Targeted Protein Degradation: Methods and Protocols [2365, 1 ed.] 1071616641, 9781071616642

Table of contents :
Preface
References
Contents
Contributors
Part I: Methods to Discover and Characterize Ligands for Targets and/or Ligases
Chapter 1: Protein Ligand Interactions Using Surface Plasmon Resonance
1 Introduction
2 Materials
2.1 SPR Assay
2.2 Compound Plate Preparation
3 Methods
3.1 Prepare the Biacore 8K+ System
3.2 Design Method in the Biacore 8K Control Software
3.3 Prepare the Assay Plates
3.4 Run the Method
3.5 Analyze the Data in Biacore Insight 8K Evaluation Software
4 Notes
References
Chapter 2: High-Throughput Detection of Ligand-Protein Binding Using a SplitLuc Cellular Thermal Shift Assay
1 Introduction
2 Materials
2.1 SDS PAGE and Western Blotting
2.2 Compounds
2.3 Constructing the SplitLuc CETSA Acceptor Plasmids
2.4 Cloning TOI into the SplitLuc CETSA Plasmid
2.5 Cell Culture and Transfection
2.6 CETSA and SplitLuc CETSA
2.7 Metal Blocks and Heating Platform for Conductive Heat Transfer (Fig. S1)
2.8 Counterscreens
3 Methods
3.1 Traditional CETSA Using Western Blot Electrophoresis to Assess the Melting Profile of the TOI
3.2 Creating a SplitLuc ``Acceptor´´ Plasmid for Rapid Cloning of SplitLuc Fusion Proteins
3.3 Cloning the TOI into a SplitLuc ``Acceptor´´ Plasmid
3.4 Transfection of HEK293T Cells for Validation of TOI-SplitLuc Expression
3.5 Low-Throughput SplitLuc CETSA (PCR Tube or 96-Well Plate)
3.6 High-Throughput SplitLuc CETSA (384-Well Format)
3.7 High-Throughput SplitLuc CETSA (1536-Well Format)
3.8 Counterscreens, Data Analysis, and Confirmation
4 Notes
References
Chapter 3: Proteins and Their Interacting Partners: An Introduction to Protein-Ligand Binding Site Prediction Methods with a F...
1 Introduction
1.1 FunFOLD3
1.2 Blind Evaluation of Methods: CASP, CAMEO and CAFA
1.3 Equation 1. Matthews Correlation Coefficient
1.4 Equation 2. Binding Site Distance Test Score
2 Materials
2.1 Software Requirements
3 Methods
3.1 Installing FunFOLD3 and Required Databases
3.2 Running FunFOLD3
3.3 Running FunFOLD3 for Function Predictions in CASP11, CASP12 and CASP13
4 Potential Outcomes
5 Notes
References
Chapter 4: Evaluating Ligands for Ubiquitin Ligases Using Affinity Beads
1 Introduction
2 Materials
2.1 Compound Syntheses
2.2 Determining the Biotin Binding Capacity of the Streptavidin Beads
2.3 E3 Pulldowns from Cell Extracts
2.4 Immunoblotting and Stained Gel Analysis
3 Methods
3.1 Synthesis of the {Biotin-CRBN Ligand} Compound (See Fig. 1a)
3.2 Synthesis of {Biotin-VHL Ligand} Compound (See Fig. 1b)
3.3 Synthesis of {Biotin-TRIM24 Ligand} Compound (See Fig. 2)
3.4 Determining the Biotin Binding Capacity of the Streptavidin Beads
3.5 E3 Pulldowns from Cell Extracts
4 Notes
References
Part II: Approaches to Study Ligase:Degrader:Target Ternary Complexes
Chapter 5: Mechanistic and Structural Features of PROTAC Ternary Complexes
1 PROTAC Ternary Complexes: Equilibria and Crystal Structures
1.1 Two-Component vs. Three-Component Binding Models
1.2 Cooperativity and the Hook Effect
1.3 Structural Basis for PROTAC Cooperativity and Selectivity
1.4 Ternary Complex Structures to Guide PROTAC Design
2 Characterisation of Ternary Complexes Using Biophysical Methods
2.1 X-Ray Crystallography
2.2 Proximity Binding Assays
2.3 Competitive and Direct Binding Assays
2.4 Isothermal Titration Calorimetry-Thermodynamics
2.4.1 Introduction to Biophysical Characterisation of PROTAC-Induced Ternary Complex Formation by Isothermal Titration Calorim...
2.4.2 Materials
2.4.3 Methods
Protein Dialysis
PROTAC Preparation
Protein Preparation
Titration Procedure
Control Titrations
Binary BET BD and Ternary Titrations
VBC Binary Titration
Data Analysis
2.5 Surface Plasmon Resonance-Kinetics
3 Cellular and Functional Characterisation of Targeted Protein Degradation Mediated Via Ternary Complex Formation
3.1 Cellular Target Engagement Assays
3.1.1 Binary Target Engagement
3.1.2 Target Engagement of the Ternary Complex
3.2 Cellular Target Ubiquitination and Degradation Assays
3.2.1 Assessing Target Ubiquitination in Cells
3.2.2 Assessing Protein Degradation in Cells
3.2.3 Assessing Protein Degradation In Vivo
4 Concluding Remarks
References
Chapter 6: MST and TRIC Technology to Reliably Study PROTAC Binary and Ternary Binding in Drug Development
1 Introduction
2 Materials
2.1 Protein Preparation
2.1.1 Proteins and Constructs
2.1.2 Protein Production and Purification
2.2 Protein Labeling
2.3 Affinity Measurements with MST
2.4 SD-Test
3 Methods
3.1 Protein Production and Purification
3.2 Protein Labeling
3.3 Binary Interaction Analysis
3.3.1 Protein-PROTAC Affinity Measurements with MST
3.3.2 Data Analysis
3.3.3 SD-Test
3.4 Ternary Interaction Analysis
3.4.1 Ternary Complex Affinity Measurements with MST
3.4.2 Data Analysis
3.4.3 Cooperativity Assessment of Ternary Complex
4 Notes
References
Chapter 7: An In Vitro Pull-down Assay of the E3 Ligase:PROTAC:Substrate Ternary Complex to Identify Effective PROTACs
1 Introduction
2 Materials
2.1 GST-VBC Protein Expression
2.2 GST-VBC Protein Purification
2.3 In Vitro Trimer Pull-down Assay (TPA)
2.4 Common Laboratory Equipment and Supplies
3 Methods
3.1 Timing
3.2 GST-VBC Purification
3.2.1 GST-VBC Protein Expression (4-Day Procedure)
3.2.2 Cell Lysis and GST Pull-down (1-Day Procedure)
3.2.3 Anion Exchange Purification (Half-Day Procedure)
3.2.4 Size Exclusion Purification (Half-Day Procedure)
3.3 In Vitro Trimer Pull-down Assay (TPA)
3.3.1 Preparing GST-VBC-Coated Beads and Dilution of the PROTAC Compounds
3.3.2 Incubation of Substrate and PROTAC with GST-VBC-Coated Beads
3.3.3 Final Washes and Elution
3.3.4 Western Blot Analysis
4 Notes
References
Chapter 8: Kinetic Detection of E3:PROTAC:Target Ternary Complexes Using NanoBRET Technology in Live Cells
1 Introduction
2 Materials
2.1 NanoBRET Ternary Complex Assays Using Example BRD4:VHL and BRD4:CRBN + BET PROTACs
3 Methods
3.1 Generation of HaloTag E3 Fluorescent Acceptor Fusion Protein and NanoLuc or HiBiT Luminescent Donor Fusion Protein
3.1.1 Use of NanoLuc Fusion Vectors as the Luminescent Donor
3.1.2 Use of a HiBiT CRISPR Cell Line as the Luminescent Donor
3.2 Bulk Transfection of HEK293 Cells
3.2.1 Transfection Conditions for NanoBRET Ternary Complex Assay with NanoLuc-BRD4 Example and HaloTag-VHL or HaloTag-CRBN or ...
3.2.2 Transfection Conditions for Tag Placement and Donor-to-Acceptor Ratio Optimization for User-Generated NanoLuc Target Fus...
3.2.3 Transfection Conditions for Assay Optimization with Endogenously Tagged HiBiT CRISPR Fusion and HaloTag E3 Fusion Vector
3.3 Replating Transfected HEK293 Cells into Multiwell Plates and Adding HaloTag NanoBRET 618 Ligand
3.4 Adding MG-132, Test Compounds, and Nano-Glo Detection Reagents
3.4.1 Live Cell Endpoint Detection Using NanoBRET Nano-Glo Detection System (Promega #N1662)
For Samples Pretreated with MG-132
For Samples NOT Pretreated with MG-132
3.4.2 Live Cell kinetic Detection Using NanoBRET Nano-Glo Kinetic Detection System (Promega #N2584)
For Samples Pretreated with MG-132
For Samples NOT Pre-Treated with MG-132
3.5 NanoBRET Calculations
4 Notes
References
Part III: Approaches and Tools to Understand Ubiquitylation and Proteasomal Degradation
Chapter 9: Inducing Ubiquitylation and Protein Degradation as a Drug Development Strategy
1 Introduction
2 Materials
2.1 Intracellular Ubiquitination
2.2 In Vitro Ubiquitination
2.3 CellTiter-Glo Luminescent Assay
2.4 Immunoprecipitation, Western Blot
3 Methods
3.1 Detecting Intracellular Protein Ubiquitination
3.2 Detecting PROTAC-Induced Protein Ubiquitination In Vitro
3.3 Determine if PROTAC-Induced Target Degradation Is UPS-Dependent
3.4 Evaluate Therapeutic Potential of Degraders by Cellular Proliferation Inhibition
3.5 Western Blot to Detect Protein Ubiquitination, Cell Apoptosis
4 Notes
References
Chapter 10: Tandem Ubiquitin Binding Entities (TUBEs) as Tools to Explore Ubiquitin-Proteasome System and PROTAC Drug Discovery
1 Introduction
2 Enrichment of Polyubiquitinated Proteins by TUBE Pulldown
2.1 Materials
2.2 Methods
2.3 Notes
3 Use of TUBEs as Detection Tools in Western Blotting
3.1 Materials
3.2 Methods
3.3 Notes
4 Use of TUBEs for Detecting Protein Polyubiquitination In Vitro
4.1 Materials
4.2 Methods
4.3 Notes
5 Use of TUBEs as Capture Reagents for Detection of Protein Polyubiquitination from Biological Matrices
5.1 Materials
5.2 Methods
5.3 Notes
6 Imaging of Polyubiquitinated Proteins Using Fluorescent-Tagged TUBEs
6.1 Materials
6.2 Methods
6.3 Notes
7 TUBEs for High-Throughput Cellular and In Vitro Assays to Monitor PROTAC Driven Protein Ubiquitination and Degradation
7.1 Materials
7.2 Methods
7.3 Notes
References
Chapter 11: Global Mass Spectrometry-Based Analysis of Protein Ubiquitination Using K-ε-GG Remnant Antibody Enrichment
1 Introduction
2 Materials
2.1 Lysis and Digestion to Peptides
2.2 Immunoaffinity Purification
2.3 C18 Sample Desalting
2.4 LC-MS/MS Analysis
3 Methods
3.1 Lysis and Digestion to Peptides
3.1.1 Cell Line Protocol
3.1.2 Tissue Protocol
3.2 Immunoaffinity Purification
3.3 C18 Sample Desalting (See Note 18)
3.4 LC-MS/MS Analysis (See Notes 20-22)
3.5 Data Analysis
4 Notes
References
Chapter 12: Determination of Proteasomal Unfolding Ability
1 Introduction
2 Materials
2.1 Protein Purification
2.2 Ubiquitination
2.3 Proteasome Purification
2.4 Unfolding Ability Assay
2.5 Data Analysis
2.6 General Equipment/Instruments
3 Methods
3.1 Substrate Preparation
3.1.1 Expression and Purification
3.1.2 Labeling (Adapted from)
3.2 Preparation of Ubiquitination Enzymes
3.2.1 Ubiquitin ([15]; Modified from)
3.2.2 E1 Enzyme (Modified from)
3.2.3 E2 Enzymes
UbcH5a and UbcH7
Ubc2
3.2.4 E3 Enzymes
Rsp5
Cul3/Rbx1 (Modified from)
Keap1 C151S (Modified from) (See Note 18)
Ubr1
3.3 Ubiquitination
3.3.1 General Protocol for ubiquitination Reactions
3.3.2 Purification of Ubiquitinated Proteins
3.4 Proteasome Purification
3.5 Unfolding Ability Assay
3.5.1 Assay
3.5.2 Gels
3.6 Data Analysis
4 Notes
References
Part IV: Methods for Evaluating Protein Degrader Function
Chapter 13: Methods for Quantitative Assessment of Protein Degradation
1 Introduction
2 Materials
2.1 Preparation of a Reporter Plasmid for Flp293T
2.2 Preparation of Reporter Line by Lentivirus
2.3 Flow Cytometry Analysis
2.4 Laser Scanning Cytometry Analysis
3 Methods
3.1 Preparation of Reporter Cell Line
3.1.1 Preparation of a Reporter Cell Line in Flp-in System
3.1.2 Preparation of a Reporter Line by Lentivirus Transduction
Lentiviral Packaging
Lentiviral Transduction
3.2 Degradation Readout
3.2.1 Flow Cytometry Analysis
Cell Seeding and Treatment
FACS Analyzer Readout
Data Analysis
3.2.2 Laser Scanning Cytometry Analysis
Cell Seeding and Treatment
Imaging Readout
Data Analysis
4 Notes
References
Chapter 14: A High-Throughput Method to Prioritize PROTAC Intracellular Target Engagement and Cell Permeability Using NanoBRET
1 Introduction
2 Materials
2.1 Transient Transfection of HEK293 Cells with NanoLuc Fusions
2.2 Determination of Compound Affinity in Live Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)
2.3 Determination of Compound Affinity in Permeabilized Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)
3 Methods
3.1 Transient Transfection of HEK293 Cells with NanoLuc Fusions
3.2 Determination of Compound Affinity in Live Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)
3.3 Determination of Compound Affinity in Permeabilized Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)
3.4 Assessment of Relative Intracellular Availability for E3 Ligase Ligands Via the Availability Index
4 Notes
References
Chapter 15: Profiling CELMoD-Mediated Degradation of Cereblon Neosubstrates
1 Introduction
1.1 Overview
1.2 High-Throughput Substrate Degradation
1.3 Chemical and Genetic Approaches to Confirm CELMoD MOA
1.4 In Vitro Ubiquitination
1.5 Cereblon Binding
2 Materials
2.1 High-Throughput Cellular Substrate Degradation
2.2 Chemical and Genetic Approaches to Confirm CELMoD MOA
2.3 In Vitro Ubiquitination
2.4 Cereblon Binding
3 Methods
3.1 High-Throughput Cellular Substrate Degradation
3.1.1 Generation of Cell Line Expressing ePL-Tagged Protein of Interest
3.1.2 Optimization of Target-Based ePL Assay
3.1.3 Cellular Dose-Response Curve Assay of ePL-Tagged Protein of Interest
3.2 Chemical and Genetic Approaches to Confirm CELMoD MOA
3.2.1 Generation of Cereblon Knockout Cells
3.2.2 Confirmation of Cereblon- and Proteasome-Dependent Regulation of Protein of Interest Via CELMoD Treatment
3.3 In Vitro Ubiquitination
3.4 Cereblon Binding
4 Notes
References
Chapter 16: Global Proteome Profiling to Assess Changes in Protein Abundance Using Isobaric Labeling and Liquid Chromatography...
1 Introduction
2 Materials
2.1 Lysis and Digestion to Peptides
2.2 Sep-Pak C18 Purification of Lysate Peptides
2.3 TMT Labeling
2.4 C18 Sample Desalting
2.5 Basic Reversed Phase (bRP) Fractionation
2.6 LC-MS/MS Analysis
3 Methods
3.1 Lysis and Digestion to Peptides
3.2 Sep-Pak C18 Purification of Lysate Peptides
3.3 TMT Labeling of Peptides
3.4 C18 Sample Desalting (See Note 10)
3.5 Basic Reversed Phase Fractionation
3.6 LC-MS/MS Analysis (See Notes 13 and 14)
3.7 Data Analysis
4 Notes
References
Chapter 17: PHOTACs Enable Optical Control of Protein Degradation
1 Introduction
2 Materials
2.1 Cell Culture
2.2 Reagents
2.3 Light Source
2.4 Immunoblotting
3 Methods
3.1 Preparation of PHOTAC Stock Solutions
3.2 PHOTAC Treatment
4 Notes
References
Chapter 18: Protocols for Synthesis of SNIPERs and the Methods to Evaluate the Anticancer Effects
1 Introduction
2 Materials
2.1 Design and Synthesis of SNIPERs
2.2 Target Protein Degradation
2.3 Cell Death
2.4 Ubiquitylation Assay
2.5 In Vivo Protein Knockdown
2.6 In Vivo Tumor Growth Inhibition
3 Methods
3.1 Design and Synthesis of SNIPER(ER)-14 and -19
3.2 Design and Synthesis of SNIPER(ER)-87
3.3 Design and Synthesis of SNIPER(ABL)-20 and -19
3.4 Design and Synthesis of SNIPER(ABL)-39
3.5 Design and Synthesis of SNIPER(BRD4)-1 and SNIPER(PDE4)-9
3.6 Analysis of Target Protein Degradation by Immunoblot
3.7 Analysis of Ubiquitylation
3.8 Analysis of Cell Death by Crystal Violet Staining
3.9 Analysis of In Vivo Protein Knockdown
3.10 Analysis of In Vivo Tumor Growth Inhibition
4 Notes
References
Index

Citation preview

Methods in Molecular Biology 2365

Angela M. Cacace Christopher M. Hickey Miklós Békés Editors

Targeted Protein Degradation Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

Targeted Protein Degradation Methods and Protocols

Edited by

Angela M. Cacace, Christopher M. Hickey, and Miklós Békés Platform Biology, Arvinas, Inc., New Haven, CT, USA

Editors Angela M. Cacace Platform Biology Arvinas, Inc. New Haven, CT, USA

Christopher M. Hickey Platform Biology Arvinas, Inc. New Haven, CT, USA

Miklo´s Be´ke´s Platform Biology Arvinas, Inc. New Haven, CT, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-1664-2 ISBN 978-1-0716-1665-9 (eBook) https://doi.org/10.1007/978-1-0716-1665-9 © Springer Science+Business Media, LLC, part of Springer Nature 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface Molecules designed to induce ubiquitylation and subsequent degradation of specific target proteins have become the focus of intense research interest across the biotech and pharmaceutical industries [1, 2]. Targeted protein degradation as a potential therapeutic strategy was first observed in the early 2000s when selective estrogen receptor degraders (SERDs) and immunomodulatory drugs (IMiDs) were shown to act by binding to and destabilizing their respective protein target causing them to be degraded by the ubiquitin-proteasome system (UPS) [3, 4]. While these early small molecules highlighted the therapeutic potential of harnessing the UPS, their usefulness was limited to the specific proteins they target. More recently, heterobifunctional molecules, known as PROTeolysis TArgeting Chimeras or PROTACs®, have been designed to specifically bind both a target protein and a ubiquitin ligase (E3), thus enabling specific ubiquitylation of the protein-of-interest and its subsequent degradation by the UPS. In principle, these bifunctional molecules can target any protein for degradation by the UPS allowing for limitless control of target selection and degradation. This flourishing field of targeted protein degradation is also stimulating basic science discoveries and providing important tools for biomedical research. Most importantly, oral PROTACs have now entered clinical development for the treatment of cancers, despite having properties that challenge traditional medicinal chemistry guidelines [2, 5]. This volume captures a collection of innovative techniques for the discovery and development of a set of revolutionary small molecules that induce ubiquitylation and subsequent degradation of target proteins via the UPS. Chapters detail examples of heterobifunctional degrader molecules (e.g., PROTACs, SNIPERs) as well as molecular glues (e.g., CELMoDs) that stabilize binding interactions between a protein and the E3. Written for the highly successful Methods in Molecular Biology series, the protocols are complemented with insightful introductory perspectives from expert authors and detailed step-by-step methods to identify and characterize new ligands; measure binary and ternary complex formation; study E3 ligase-induced ubiquitylation; examine proteasomal degradation of proteins; apply proteomic approaches to understand selectivity of targeted protein degraders; and discover novel degrader molecules. Authoritative and cutting-edge, Targeted Protein Degradation: Methods and Protocols aims to ensure successful results in this emerging field of drug discovery. Special and sincere thanks to all contributors who made this comprehensive volume possible. New Haven, CT, USA

Angela M. Cacace Miklos Be´ke´s Christopher M. Hickey

References 1. Sakamoto KM, Kim KB, Kumagai A et al (2001) Protacs: chimeric molecules that target proteins to the Skp1-Cullin-F box complex for ubiquitination and degradation. Proc Natl Acad Sci USA 98:8554–8559

v

vi

Preface

2. Pettersson M, Crews CM (2019) PROteolysis Targeting Chimeras (PROTACs)—past, present and future. Drug Discov Today Technol 31:15–27 3. Cornella-Taracido I, Garcia-Echeverria C (2020) Monovalent protein-degraders—insights and future perspectives. Bioorg Med Chem Lett 30:127202 4. Lu G, Middleton RE, Sun H et al (2014) The myeloma drug lenalidomide promotes the cereblondependent destruction of Ikaros proteins. Science 343:305–309 5. Cantrill C, Chaturvedi P, Rynn C et al (2020) Fundamental aspects of DMPK optimization of targeted protein degraders. Drug Discov Today 25:969–982

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

METHODS TO DISCOVER AND CHARACTERIZE LIGANDS FOR TARGETS AND/OR LIGASES

1 Protein Ligand Interactions Using Surface Plasmon Resonance . . . . . . . . . . . . . . . Nichole O’Connell 2 High-Throughput Detection of Ligand-Protein Binding Using a SplitLuc Cellular Thermal Shift Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tino W. Sanchez, Ashley Owens, Natalia J. Martinez, Eric Wallgren, Anton Simeonov, and Mark J. Henderson 3 Proteins and Their Interacting Partners: An Introduction to Protein–Ligand Binding Site Prediction Methods with a Focus on FunFOLD3 . . . . . . . . . . . . . . . Danielle Allison Brackenridge and Liam James McGuffin 4 Evaluating Ligands for Ubiquitin Ligases Using Affinity Beads . . . . . . . . . . . . . . . Jennifer Dobrodziej, Hanqing Dong, Kurt Zimmermann, and Christopher M. Hickey

PART II

v ix

3

21

43 59

APPROACHES TO STUDY LIGASE:DEGRADER:TARGET TERNARY COMPLEXES

5 Mechanistic and Structural Features of PROTAC Ternary Complexes . . . . . . . . . 79 Ryan Casement, Adam Bond, Conner Craigon, and Alessio Ciulli 6 MST and TRIC Technology to Reliably Study PROTAC Binary and Ternary Binding in Drug Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Tanja Bartoschik, Andreas Zoephel, Klaus Rumpel, Alessio Ciulli, and Charles Heffern 7 An In Vitro Pull-down Assay of the E3 Ligase:PROTAC:Substrate Ternary Complex to Identify Effective PROTACs. . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Daniel P. Bondeson, Blake E. Smith, and Alexandru D. Buhimschi 8 Kinetic Detection of E3:PROTAC:Target Ternary Complexes Using NanoBRET Technology in Live Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Sarah D. Mahan, Kristin M. Riching, Marjeta Urh, and Danette L. Daniels

PART III

APPROACHES AND TOOLS TO UNDERSTAND UBIQUITYLATION AND PROTEASOMAL DEGRADATION

9 Inducing Ubiquitylation and Protein Degradation as a Drug Development Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Kamaldeep S. Dhami and XiaoDong Huang

vii

viii

10

11

12

Contents

Tandem Ubiquitin Binding Entities (TUBEs) as Tools to Explore Ubiquitin-Proteasome System and PROTAC Drug Discovery. . . . . . . . . . . . . . . . 185 Karteek Kadimisetty, Katie J. Sheets, Patrick H. Gross, Myra J. Zerr, and Dahmane Ouazia Global Mass Spectrometry-Based Analysis of Protein Ubiquitination Using K-ε-GG Remnant Antibody Enrichment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Alissa J. Nelson, Yiying Zhu, Jian Min Ren, and Matthew P. Stokes Determination of Proteasomal Unfolding Ability . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Christina M. Hurley and Daniel A. Kraut

PART IV 13 14

15

16

17 18

METHODS FOR EVALUATING PROTEIN DEGRADER FUNCTION

Methods for Quantitative Assessment of Protein Degradation . . . . . . . . . . . . . . . . Radosław P. Nowak, Hong Yue, Emily Y. Park, and Eric S. Fischer A High-Throughput Method to Prioritize PROTAC Intracellular Target Engagement and Cell Permeability Using NanoBRET . . . . . . . . . . . . . . . . James D. Vasta, Cesear R. Corona, and Matthew B. Robers Profiling CELMoD-Mediated Degradation of Cereblon Neosubstrates. . . . . . . . Joel W. Thompson, Thomas Clayton, Gody Khambatta, Leslie A. Bateman, Christopher W. Carroll, Philip P. Chamberlain, and Mary E. Matyskiela Global Proteome Profiling to Assess Changes in Protein Abundance Using Isobaric Labeling and Liquid Chromatography-Tandem Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anthony P. Possemato, Kathryn Abell, and Matthew P. Stokes PHOTACs Enable Optical Control of Protein Degradation . . . . . . . . . . . . . . . . . . Martin Reynders and Dirk Trauner Protocols for Synthesis of SNIPERs and the Methods to Evaluate the Anticancer Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yoshinori Tsukumo, Genichiro Tsuji, Hidetomo Yokoo, Norihito Shibata, Nobumichi Ohoka, Yosuke Demizu, and Mikihiko Naito

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

247

265 283

301 315

331

349

Contributors KATHRYN ABELL • Cell Signaling Technology, INC., Danvers, MA, USA TANJA BARTOSCHIK • NanoTemper Technologies GmbH, Munich, Germany LESLIE A. BATEMAN • Bristol Myers Squibb Company, San Diego, CA, USA ADAM BOND • Division of Biological Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, Dundee, Scotland, UK DANIEL P. BONDESON • The Broad Institute of MIT and Harvard, Cambridge, MA, USA DANIELLE ALLISON BRACKENRIDGE • School of Biological Sciences, University of Reading, Reading, UK ALEXANDRU D. BUHIMSCHI • Northwestern University Feinberg School of Medicine, Chicago, IL, USA CHRISTOPHER W. CARROLL • Bristol Myers Squibb Company, San Diego, CA, USA RYAN CASEMENT • Division of Biological Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, Dundee, Scotland, UK PHILIP P. CHAMBERLAIN • Bristol Myers Squibb Company, San Diego, CA, USA ALESSIO CIULLI • Division of Biological Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, Dundee, Scotland, UK; School of Life Sciences, University of Dundee, Scotland, UK THOMAS CLAYTON • Bristol Myers Squibb Company, San Diego, CA, USA CESEAR R. CORONA • Promega Corporation, Fitchburg, WI, USA CONNER CRAIGON • Division of Biological Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, Dundee, Scotland, UK DANETTE L. DANIELS • Promega Corporation, Fitchburg, WI, USA YOSUKE DEMIZU • Division of Organic Chemistry, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan KAMALDEEP S. DHAMI • Oncology Discovery, AbbVie, Sunnyvale, CA, USA JENNIFER DOBRODZIEJ • Arvinas, Inc., New Haven, CT, USA HANQING DONG • Arvinas, Inc., New Haven, CT, USA ERIC S. FISCHER • Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA, USA; Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA, USA PATRICK H. GROSS • Department of Research and Development, LifeSensors, Inc., Malvern, PA, USA CHARLES HEFFERN • NanoTemper Technologies GmbH, Munich, Germany MARK J. HENDERSON • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA CHRISTOPHER M. HICKEY • Platform Biology, Arvinas, Inc., New Haven, CT, USA XIAODONG HUANG • Oncology Discovery, AbbVie, Sunnyvale, CA, USA CHRISTINA M. HURLEY • Department of Chemistry, Villanova University, Villanova, PA, USA KARTEEK KADIMISETTY • Department of Research and Development, LifeSensors, Inc., Malvern, PA, USA GODY KHAMBATTA • Bristol Myers Squibb Company, San Diego, CA, USA DANIEL A. KRAUT • Department of Chemistry, Villanova University, Villanova, PA, USA

ix

x

Contributors

SARAH D. MAHAN • Promega Corporation, Fitchburg, WI, USA NATALIA J. MARTINEZ • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA MARY E. MATYSKIELA • Bristol Myers Squibb Company, San Diego, CA, USA LIAM JAMES MCGUFFIN • School of Biological Sciences, University of Reading, Reading, UK MIKIHIKO NAITO • Divisions of Molecular Target and Gene Therapy Products, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan ALISSA J. NELSON • Cell Signaling Technology, INC, Danvers, MA, USA RADOSŁAW P. NOWAK • Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA, USA; Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA, USA NICHOLE O’CONNELL • Kymera Therapeutics, Watertown, MA, USA NOBUMICHI OHOKA • Divisions of Molecular Target and Gene Therapy Products, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan DAHMANE OUAZIA • Department of Business Development, LifeSensors, Inc., Malvern, PA, USA ASHLEY OWENS • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA EMILY Y. PARK • Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA, USA ANTHONY P. POSSEMATO • Cell Signaling Technology, INC., Danvers, MA, USA JIAN MIN REN • Cell Signaling Technology, INC, Danvers, MA, USA MARTIN REYNDERS • Department of Chemistry, New York University, New York, NY, USA; Department of Chemistry, Ludwig Maximilians University of Munich, Munich, Germany KRISTIN M. RICHING • Promega Corporation, Fitchburg, WI, USA MATTHEW B. ROBERS • Promega Corporation, Fitchburg, WI, USA KLAUS RUMPEL • Boehringer Ingelheim RCV GmbH & Co KG, Vienna, Austria TINO W. SANCHEZ • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA KATIE J. SHEETS • Department of Research and Development, LifeSensors, Inc., Malvern, PA, USA NORIHITO SHIBATA • Divisions of Molecular Target and Gene Therapy Products, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan ANTON SIMEONOV • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA BLAKE E. SMITH • Harvard Medical School, Boston, MA, USA MATTHEW P. STOKES • Cell Signaling Technology, INC., Danvers, MA, USA JOEL W. THOMPSON • Bristol Myers Squibb Company, San Diego, CA, USA DIRK TRAUNER • Department of Chemistry, New York University, New York, NY, USA; Perlmutter Cancer Center, New York University School of Medicine, New York, NY, USA; NYU Neuroscience Institute, New York University School of Medicine, New York, NY, USA GENICHIRO TSUJI • Division of Organic Chemistry, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan YOSHINORI TSUKUMO • Divisions of Molecular Target and Gene Therapy Products, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan MARJETA URH • Promega Corporation, Fitchburg, WI, USA JAMES D. VASTA • Promega Corporation, Fitchburg, WI, USA

Contributors

xi

ERIC WALLGREN • National Center for Advancing Translational Sciences, National Institutes of Health, Rockville, MD, USA HIDETOMO YOKOO • Organic Chemistry, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan HONG YUE • Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA, USA; Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA, USA MYRA J. ZERR • Department of Business Development, LifeSensors, Inc., Malvern, PA, USA YIYING ZHU • Cell Signaling Technology, INC, Danvers, MA, USA KURT ZIMMERMANN • Arvinas, Inc., New Haven, CT, USA ANDREAS ZOEPHEL • Boehringer Ingelheim RCV GmbH & Co KG, Vienna, Austria

Part I Methods to Discover and Characterize Ligands for Targets and/or Ligases

Chapter 1 Protein Ligand Interactions Using Surface Plasmon Resonance Nichole O’Connell Abstract Surface Plasmon Resonance (SPR) is a powerful biophysical method for characterizing small molecule binding to proteins. Owing to its ability to characterize binary inteactions between warheads and E3 ligases or substrates, SPR is a useful tool for the development of targeted protein degraders. SPR is also an effective method for optimizing linkers and characterizing ternary complex interactions that are mediated by heterobifunctional ligands (Roy et al. ACS Chem Biol 14:361–368, 2019). Recent advances in the throughput of modern instruments have improved the ability of SPR to rapidly triage ligands based on binding kinetics and affinity, making this technique invaluable for driving degrader optimization. This chapter describes the characterization of ligands binding to the Thalidomide Binding Domain of mouse Cereblon (mCRBN-TBD) using the Biacore 8K+. Key words SPR, Biacore, Binding affinity, Binding kinetics, Cereblon, Thalidomide, Pomalidomide, Lenalidomide, ARV-825, dBET-1

1

Introduction A critical first step in the development of potent heterobifunctional degraders is the optimization of binary interactions between small molecule warheads and their respective E3 and substrate. The mechanism of action of heterobifunctional degraders is to bridge non-native protein-protein interactions. As such, the criteria for a suitable ligand requires a potent binding interaction rather than a potent inhibition of the target. For E3 ligase ligand discovery, inhibition of the protein would render the enzyme incapable of ubiquitination and is highly undesirable. For the substrate warhead, inhibition is tolerated but not required and may give rise to off-target interactions due to the poor selectivity, as is the case for many kinase inhibitors (see Note 1 for special considerations) [1]. It can be advantageous for these binding interactions to be functionally silent. In this case, classic biochemical assays are unsuitable for driving compound development as they are unable to detect silent

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

3

4

Nichole O’Connell

Fig. 1 The Physics of SPR. The resonance condition, θresonance, is dictated by the refractive index at the sensor surface. Incident light that satisfies the resonance condition is absorbed leaving a shadow on the detector. As compounds bind, the refractive index at the sensor surface changes and the shadow moves. The green arrow highlights the movement of the shadow upon compound binding. This movement over the time defines the SPR sensorgram, as described in Fig. 2

binders. Biophysical methods such as SPR are well-suited to detect binding interactions at any site on the protein and have therefore become a valuable tool in the development of heterobifunctional degraders for therapeutic use [2–4]. SPR is a biophysical technique with exquisite sensitivity to detect small changes in mass captured at the surface of a sensor chip. The sensor is comprised of a glass slide coated with a thin layer of gold which is coupled to an optical interface (see Fig. 1). The gold provides a platform to immobilize bioactive molecules (ligands), and serves as a medium for the generation of a surface plasmon which enables detection of subtle variations in refractive index at the surface. The bioactive surface sits inside a fluidic cell through which small molecules (analytes) can be introduced under constant flow for a discrete amount of time. Compound binding causes a change in mass density and thus the refractive index at the surface, which is detected by the optical interface, leading to an increase in the signal response (see Note 2) [5]. A basic SPR experiment is depicted in Fig. 2. First, the baseline signal is established under the buffer flow and then a small molecule analyte is injected. Finally, buffer flow is restored to allow the small molecule to dissociate from the protein. The resulting signal (sensorgram) consists of 3 phases: the association phase, steady-state equilibrium, and the dissociation phase. In a typical multi-cycle experiment, sensorgrams are collected over an analyte titration and used to fit the data to obtain information about the equilibrium

Protein Ligand Interactions by SPR

5

Fig. 2 A Basic SPR Experiment. (1) The baseline signal is established under buffer flow. (2) Compound is injected. Binding causes a rise in the signal response during the association phase. (3) Steady state equilibrium is reached and the signal no longer changes as the number of molecules binding equals the number of molecules dissociating from the surface. (4) Buffer flow is restored to wash the molecules away from the surface during the dissociation phase. (5) In a multi-cycle experiment, the process is repeated over a compound titration. (6) The data is analyzed to determine kinetic parameters (when possible) and steady state affinity as described in Note 3

6

Nichole O’Connell

Fig. 3 Ligand and analytes. The structure of mCRBN-TBD bound to thalidomide is shown in purple (Left panel) [8]. Three warheads and two heterobifunctional molecules were evaluated (Right panel). dBET1 links Thalidomide and ARV-825 links Pomalidomide to the BRD4 binder, JQ1

dissociation constant (KD) and binding kinetics (kon, koff) under the right conditions [6]. Two methods are commonly employed to analyze the data, kinetic and steady state, as described in Note 3. SPR experiments allow for rapid triaging of compound binding, enabling subsequent design optimization based on affinity and kinetic parameters. This protocol describes the use of SPR to characterize binary interactions between mCRBN-TDB [7] and IMiD molecules thalidomide, pomalidomide and lenalidomide as well as the heterobifunctional degraders dBET1 and ARV-825 that target BRD4 for degradation (Fig. 3) [9–11]. This protocol is specific for screening compounds using the state-of-the-art, high throughput 8K+ instrument from Biacore.

2 2.1

Materials SPR Assay

1. Biacore 8K+ (Cytiva 29283382). 2. Series S Maintenance Chip (Cytiva BR100562). 3. Desorb Kit (Cytiva BR100823): BIAdesorb solution 1 and BIAdesorb solution 2. 4. Series S SA chip (Cytiva 29104992). 5. HBS-EP+ Buffer 10
: 0.1 M HEPES, 1.5 M NaCl, 0.03 M EDTA and 0.5% v/v Surfactant P20, pH 7.4 (Cytiva BR100669).

Protein Ligand Interactions by SPR

7

6. Bond Breaker TCEP Solution: 0.5 M TCEP (ThermoFisher 77720). 7. 96 well polypropylene plate, 1 mL (Greiner Bio-one 780210). 8. 384 deep well microplate (Greiner Bio-one 781270). 9. Microplate foil (384 well) (Cytiva BR100577). 10. 96-well septa (Cytiva 29192561). 11. 0.2 μm PES sterile filter top, 500 mL, Corning 431118. 12. 5 M NaCl. 13. 0.1 M NaOH. 14. Isopropanol. 15. Biotinylated Mouse Cereblon Thalidomide Binding Domain (mCRBN-TBD): SGS-CRBN-TBD (318-427)-Avi Tag (prepared as described in [7]). 2.2 Compound Plate Preparation

1. Dimethyl sulfoxide (Sigma Aldrich D8418-1L). 2. Thalidomide (Selleckchem S1193). 3. Pomalidomide (Selleckchem S1567). 4. Lenalidomide (Selleckchem S1029). 5. dBET1 (Selleckchem S8296). 6. ARV-825 (Selleckchem S8297). 7. Tecan d300e Digital Printer. 8. T8 Plus Dispense Cassette (Tecan 30097370). 9. D4 Plus Dispense Cassette (Tecan 30097371). 10. MultiDrop Combi (Thermo Scientific 5840300). 11. MultiDrop Combi Standard Tube Dispensing Cassette (Thermo Scientific 24072670).

3

Methods

3.1 Prepare the Biacore 8K+ System

1. Remove a streptavidin coated SA sensor chip from storage at 4  C and allow it to equilibrate to room temperature for 15–30 min. 2. Dock a maintenance chip into the Biacore 8K+ using the CHANGE CHIP command. Using a 96 well 1 mL deep-well block, fill column 1 with 1 mL/well BIAdesorb solution 1 and column 2 with 1 mL/well BIAdesorb solution 2. Run the desorb protocol. 3. Prepare assay buffer. Special care must be taken to avoid diluting the DMSO concentration of the buffer. First, prepare 2 L of buffer without DMSO by mixing 200 mL 10
HBS-EP+ with 4 mL 0.5 M BondBreaker TCEP Solution. Bring the volume

8

Nichole O’Connell

up to 2 L using Milli-Q water. Filter the buffer into a 2 L bottle using a vacuum driven sterile filter top. Remove 200 mL buffer without DMSO and set aside for preparation of the assay plates. Remove an additional 18 mL of buffer and then add 18 mL 100% DMSO to create 1800 mL assay buffer at precisely 1% DMSO. Set aside some buffer with DMSO for startup and negative control samples. 4. Insert the buffer line of the 8K+ into the assay buffer. Place the water and reagent line in Milli-Q water. Prime all lines of the 8K+ twice using the CHANGE SOLUTIONS command. 5. Eject the maintenance chip and dock the SA chip using the CHANGE CHIP command. 6. Prime the system twice as in 4, except this time with the reagent line inserted in the reagent bottle (see Note 4). 7. Let the system sit in standby for at least 1 h while you prepare the assay plates. 3.2 Design Method in the Biacore 8K Control Software

Before setting up the assay plates, the experiment was designed using Method Builder in the Biacore 8K Control Software to allow the assay plates to be properly mapped to the experimental design. 1. Open the Method tab in the Control Software. Select New and then Empty method and thereafter EMPTY ANALYSIS METHOD. First, set up the method architecture using the tab 1. METHOD DEFINITION. This method consists of 5 steps: Conditioning, Startup, Capture, Startup 2 and Analysis. Each step is defined by a name and a purpose. Commands for each step are added using the ADD COMMAND button. All injections are run using high performance injections. (a) Conditioning: Add a step named CONDITIONING with purpose, CONDITIONING, to condition both flow cells according to the manufacturer’s recommendation. Set up an injection by adding the command ANALYTE to inject the solution 1 M NaCl/50 mM NaOH over flow cells (FC) 1 and 2 for a contact time of 60 s, dissociation time of 60 s at flow rate of 30 μL/min. Add a WASH command to wash the flow system with freshly prepared 1 M NaCl/ 50 mM NaOH/50% isopropanol (see Note 5). (b) Startup: Add a step named STARTUP with purpose, STARTUP, to equilibrate the surface prior to immobilization. Add command ANALYTE to inject assay buffer over both flow cells for 60 s contact time, 60 s dissociation time at 30 μL/min. Add command WASH to wash the system with 50% DMSO.

Protein Ligand Interactions by SPR

9

(c) Capture: Add a General step named CAPTURE with purpose, GENERAL, to immobilize the protein of interest. Define this CAPTURE step to inject only onto FC2 by selecting this flow path. Define the injection solution to be 0.5 μM biotinylated-mCRBN TBD. Set the contact time to 40 s and flow rate to 10 μL/min. Add a WAIT command with waiting time of 3600 s using a custom flow rate of 10 μL/min to give the protein time to equilibrate on the surface (see Note 6). (d) Startup 2: Add a step named STARTUP 2 with purpose, STARTUP, in order to equilibrate the surface prior to compound screening. Add command ANALYTE to inject assay buffer over both flow cells for 60 s contact time, 60 s dissociation time at 30 μL/min. Add command WASH to wash the system with 50% DMSO. (e) Analysis: This step defines the parameters for all compound injections. Add a step named ANALYSIS with the purpose, ANALYSIS. Add command ANALYTE to set up compound injections with 60 s contact time, 300 s dissociation time at a flow rate of 50 μL/min (see Note 7). Make sure the solution property is selected to be a variable. Add command WASH to wash the system with 50% DMSO (see Note 8). Within the analysis step, there are nested steps that repeat with a defined periodicity, namely Positive Control (3 μM Pomalidomide, see Note 9), Negative Control (assay buffer, see Note 10), and Solvent Correction (four samples of varying DMSO concentration, see Note 11). Add a step named POSITIVE CONTROL with purpose, ANALYSIS. Check the box REPEAT WITHIN to nest this step within the Analysis step. Select RUN ONCE FIRST, RUN ONCE LAST, and to run every 12 cycles. This will add a control injection between each compound titration. Add command ANALYTE to set up injections with 30 s contact time, 120 s dissociation time at a flow rate of 50 μL/min. Define the Positive Control solution, concentration, and molecular weight as Pomalidomide, 3 μM and 273 Da, respectively. Add command WASH to wash the system with 50% DMSO. Repeat this process to add a nested step named NEGATIVE CONTROL to run with the same periodicity as the positive control. In the analyte command, name the solution BUFFER and set the concentration and molecular weight to 0. Finally, add a step named SOLVENT CORRECTION with the purpose of solvent correction. Click the box REPEAT WITHIN to nest this step within the ANALYSIS step. Select RUN ONCE FIRST and RUN ONCE LAST and select to run every 400 cycles (see Note 12).

10

Nichole O’Connell

2. Navigate to tab 2. VARIABLES AND POSITIONING. This section defines the number of channels used, number of cycles for each step, plate positions for all solutions, and any variables associated with the steps defined in Method definition. In this experiment, four channels are used. Select USE CHANNELS 1–4. Under Plate 1, select 384 DEEP-WELL 200 μL. Under Plate 2 select 96 DEEP-WELL 1000 μL (see Note 13). To minimize the number of wells used in the experiment, pooling of solutions that are used repeatedly was enabled by clicking the gear button to adjust POSITION SETTINGS and selecting POOLING for the appropriate solutions. (a) Conditioning: Add 3 cycles. In this experiment, the preconditioning solution, and preconditioning wash solution were positioned in column 3 and 4 of Plate 2, respectively (see Note 14). (b) Startup: Add 5 cycles. Assay buffer for Startup injections was positioned in column 1 of Plate 2. (c) Capture: Run 1 cycle. mCRBN-TBD at 0.5 μM was positioned in column 9 of Plate 2. (d) Startup 2: Add 10 cycles. Assay buffer for Startup 2 injections was positioned in column 2 of Plate 2. (e) Analysis: To define the compound cycles, a compound plate map was defined in Microsoft Excel. The spreadsheet consisted of fields for compound name, concentration, well position, plate position, and molecular weight. The plate map information was copied to the clipboard from the spreadsheet and imported using the IMPORT FROM CLIPBOARD function. Four compounds are injected per cycle. The first 12 cycles pull samples from rows A, C, E, and G of the compound plate, which correspond to Thalidomide, Pomalidomide, Lenalidomide, and ARV-825, respectively. Samples are screened from column 1 to 12 and are ordered from low to high concentration in the 384-well plate. Columns 1 and 2 contain blank samples for double referencing (see Note 17). Columns 3–12 contain compound samples for a 10-point dose response from 0.195 to 100 μM following a two-fold dilution scheme, with column 12 containing the highest concentration. Cycles 13–24 pull samples from rows B, D, F, and H. dBET1 was dosed in row B (wells 1–12). The remaining rows were filled with assay buffer. Preparation of the compound plate is described below. The 50% DMSO wash solution was positioned in the reagent bottle. Solvent correction samples were positioned in columns 5–8 of the 96-well plate from low to high. Negative and Positive control samples were positioned in columns 11 and 12 of the 96-well plate, respectively.

Protein Ligand Interactions by SPR

11

3. Next, review tab 3. CYCLE OVERVIEW which lays out all cycles for the entire experiment. Make sure it matches expectations, particularly with respect to position of nested cycles within Analysis. 4. Finally, review the 4. PLATE LAYOUT. This tab gives the user an overview of each solution needed and their respective plate positions. Use this information to guide the assay plate setup (see Note 15). 3.3 Prepare the Assay Plates

1. Compound plate preparation: To prepare the compound plate, it is recommended to dose the plate with compounds dissolved in DMSO. The volume of the compound drop will supply the sample with the DMSO required to match the 1% DMSO concentration in the assay buffer. In this experiment, the well volume is 200 μL. Thus, 2 μL drops of the compounds are dosed at a starting concentration 100-fold above the final concentration. Plate the compounds into a Greiner 384 deep well plate using the Tecan d300e Digital Printer to create the dose response and normalize all wells to a volume of 2 μL. Dose blank wells with 2 μL of neat DMSO. Next, use the MultiDrop Combi to deliver 150 μL of Assay Buffer without DMSO to each well. Manually pipette the remaining 48 μL of Assay Buffer without DMSO into each well. Cover the plate with a 384 well microplate foil (see Note 16). Shake the plate at 950 RPM for 30 min while preparing the support plate. 2. Support plate preparation: Using the volume summary as a guide (see Note 15), prepare all solutions including preconditioning solution (1 M NaCl/50 mM NaOH), preconditioning wash solution (1 M NaCl/50 mM NaOH/50% isopropanol), solvent correction solutions 1–4 (see Note 11), and positive control (3 μM Pomalidomide). Use the Assay Buffer with DMSO for startup and negative control samples. Fill the plate according to the plate map defined in the Control Software.

3.4

Run the Method

3.5 Analyze the Data in Biacore Insight 8K Evaluation Software

In Method Builder 4. PLATE LAYOUT click SEND TO QUEUE. This takes you to the instrument control tab of the Control Software. Open the hotel door and remove a tray. Place the plates securely in the tray and return them to the top position in the hotel. Click READY TO START and define the result file location, usually stored in the same directory as the method. 1. Open the result file with a defined Analysis Method: Open Insight and log into the database. Select CREATE NEW EVAL UATION, then choose the experimental result from the directory in the left pane. Click SELECT EVALUATION METHOD and choose PREDEFINED. For this experiment, a kinetics/affinity method is appropriate. Choose the LMW

12

Nichole O’Connell

folder from the list pane then select LMW MULTI-CYCLE AFFINITY— EVALUATION METHOD to open the data. This method applies a solvent correction and creates four tabs, QC—SENSORGRAM, QC—BASELINE, QC—BINDING TO REFERENCE, and EVALUATION—AFFINITY (see Note 17). 2. Apply solvent correction: The software automatically opens the solvent correction tool to correct the data. Visually inspect the sensorgrams and apply the correction (see Note 18). 3. Check data quality: (a) QC—Sensorgram: Visually inspect the sensorgrams by taking note of any anomalous curves or data that falls outside the solvent correction range (see Note 19). (b) QC—Baseline: This panel plots the value of the baseline report point (see Fig. 1) as a function of cycle. Check that the baseline is stable throughout the course of the experiment (see Note 20). (c) QC—Binding to reference: This panel plots the value of the report point ANALYTE STABILITY EARLY for the reference flow cell vs. cycle number. Make note of the cycles that have large reference binding and omit the curve from data analysis if necessary (see Note 21). (d) Control binding: Additional QC plots are helpful to understand the performance of the surface throughout the experiment, such as control binding (see Notes 9 and 10). Create a control plot by clicking the HOME tab to create a plot under NEW EVALUATION ITEMS. The default plot displays the ANALYTE BINDING LATE report point (taken 5 s before the end of the injection) vs.cycle. To display only the control data, deselect all the items except the controls under ANALYSIS STEP NAME in the SELECT SENSORGRAMS panel on the left of the screen. The positive control signal in this experiment decays over time, but this is not due to the loss of surface activity (see Note 22). Note any positive control cycles that are significantly perturbed as they likely indicate that the preceding compound was poorly behaved. If signal for the subsequent controls is also perturbed, you can infer that the bad actor damaged the chip surface and all data following this compound must be excluded from analysis. (e) Capture level: It is necessary to create a plot to display the amount of protein immobilized to the surface. Use this value, Rimmobilized, to calculate Rmax_theory as described in Eq. (5) for each compound. Create a plot from the Home menu as described in the previous step. In the SELECT

Protein Ligand Interactions by SPR

13

SENSORGRAMS panel, select GENERAL under ANALYSIS STEP PURPOSE and change the sensorgram type to ACTIVE, since the target is only immobilized to the active flow cell. Make sure the y-axis settings are set to Report point ¼ Capture level_1 and Response type ¼ Relative. The immobilization level can be reported in the table at the bottom of the screen by clicking the gear button to the right of the panel and adding Capture level_1 under REPORT POINTS RELATIVE. 4. Analyze the multi-cycle data (see Note 23): (a) Inspect each multi-cycle data set. If there is systemic artifact or significant drift in the baseline, check the blanks, as they are often the cause, and remove anomalous blank cycles by selecting the curve, right-clicking and choosing EXCLUDE LOCAL. (b) Remove traces from the multi-cycle dose response curves that have reference binding or artifact by selecting the curve, right-clicking and choosing EXCLUDE LOCAL. This is common for data collected at high compound concentration, since SPR is highly sensitive to compound insolubility. For all compounds, the 100 μM data was removed because the signal did not return to baseline after the injection. For Pomalidomide and dBET-1, the 50 μM point was also removed due to failure to restore baseline. For Thalidomide, the 3.125 μM point was removed due to artifact. For ARV-825, all data at concentrations greater than 6.25 μM were removed due to suspected insolubility. (c) Determine the appropriate fit model. For compounds that reach equilibrium at all concentrations (Fig. 2), particularly those with rectangular sensorgrams, a steady state affinity model is most appropriate as shown for Thalidomide, Pomalidomide, Lenalidomide, dBET-1 and ARV-825 (Fig. 4). In the FIT MODELS panel under Settings, choose the appropriate position to plot the response vs. concentration by changing the value under CALCULATE RESPONSE AT POSITION. For dBET1 and ARV-825, the default position of 5 s before injection end was used. However, for Thalidomide, Pomalidomide, and Lenalidomide, some non-specific binding is observed at this time point, so the report point position was moved to 5 s after injection start. Additionally, the data for ARV-825 has sufficient curvature to be fit by a kinetic fit model. Under KINETICS/AFFINITY MODE select BOTH to apply both fit models to the data. Fit results are summarized in Fig. 4.

14

Nichole O’Connell

RU vs. Conc

Sensorgram

Fit Parameters

Response (RU)

Thalidomide Steady State Affinity

24

20

16

15

8

10

KD ( M)

16.3 ± 1.3

Rmax (RU)

23.4 ± 0.6

0

5 0 1E-7

-8 1E-4

-80

0

80

160

Response (RU)

Pomalidomide

24 16 8 0 1E-7

Steady State Affinity

40 30 20 10 0 -10

32

-80

1E-4

0

80

KD ( M)

4.14 ± 0.2

Rmax (RU)

34.2 ± 0.4

160

Response (RU)

Lenalidomide

24 16 8 0 1E-7

Steady State Affinity

40 30 20 10 0 -10

32

1E-4

-80

0

80

KD ( M)

4.62 ± 0.2

Rmax (RU)

32.6 ± 0.2

160

Response (RU)

dBET1

120 80 40 0 1E-7

Steady State Affinity

160 120 80 40 0 -40

160

1E-4

-80

0

80

KD ( M)

3.63 ± 0.7

Rmax (RU)

137.1 ± 5.8

160

ARV-825 Response (RU)

120 80

Steady State Affinity 1.63 ± 0.3 KD ( M) Rmax (RU) 123.1 ± 4.7

120 80 40

40 0 1E-7 1E-5 Concentration (logM)

0 -40 -80

0 80 Time (s)

160

Kinetic 1:1 Binding kon (M-1s-1) 1.0E4 ± 2.1E3 koff (M-1) 1.5E-3 ± 1E-6 KD ( M) 1.01 Rmax (RU) 118.6 ± 0.8

Fig. 4 Summary of Results. Plots of the response vs. concentration and the sensorgrams are shown for each compound together with the fit parameters derived from analysis of the data

Protein Ligand Interactions by SPR

15

(d) Assess the quality of the fits by visual inspection to make sure they appropriately describe the data. Compare the fit Rmax to the Rmax_theory. Review the fit statistics in the Results table. Click the gear button to the right of the Results table to add standard errors (SE) and the U-value to the table. Select COLUMNS. Under AVAILABLE COLUMNS navigate to STEADY STATE AFFINITY and add SE(KD) and SE(Rmax) to the list of Selected columns. Under 1:1 BINDING add SE(ka), SE(kd), SE (Rmax) and U-value to the list of Selected columns (see Note 24). Review the standard errors for all fit parameters to verify that the parameters are statistically significant. Check the quality of the fit as measured by the chi-squared value. Steady state fits are most reliable if there are at least two data points at concentrations above the KD, even if the chi-squared is low. Additionally, for kinetic fits, look at the magnitude of the residuals and avoid fits that have large or systematic residuals. Monitor the uniqueness of the fit given by the U-value. Low U-values indicate that there is no correlation between the kinetic rate constants and the parameters are reliable. Avoid U-values above 25. For more information on fitting, please refer to [12].

4

Notes 1. As part of a heterobifunctional molecule, promiscuous kinase inhibitors have been shown to have increased degradation selectivity stemming from specific protein-protein interactions that are formed between the substrate and ligase [13]. 2. Surface plasmon resonance occurs when incident light satisfies a critical angular relationship to the sensor surface, θresonance (Fig. 1). This causes a loss in the intensity of the light which results in a shadow on the detector. As the refractive index at the sensor surface changes due to compound binding, the angle of light that satisfies the resonance condition changes and the shadow moves. This shadow movement is what defines the SPR sensorgram which is measured in response units (RU). One RU is defined by an angular change of 0.0001 which roughly correlates to 1 pg/mm2 change in mass density [6, 12]. 3. In steady state affinity analysis, the response is taken at a fixed point in time for each concentration screened (typically 5 s before the end of the injection at the ANALYTE BINDING LATE report point, Fig. 2) and plotted versus concentration.

16

Nichole O’Connell

The KD is determined by a non-linear least square fit of the data to the equation: RðC Þ ¼

Rmax C KD þ C

ð1Þ

Rmax is the maximum binding signal at saturation, KD is the equilibrium dissociation constant and C is the compound concentration. Note that this fit method is valid only when applied to data that has reached steady state equilibrium and the sensorgram has plateaued. To obtain kinetic parameters, a global fit is performed on the multi-cycle data to find the onand off-rates that best describe the dataset as a whole. To do so, the association and dissociation phases of the sensorgram are fit using the following equations: Association:   k R C Rðt Þ ¼ on max ð2Þ 1  eðkon Cþkoff Þt kon C þ koff Dissociation: Rðt Þ ¼ Rmax ekoff ðtt 0 Þ

ð3Þ

Here, kon is the forward binding rate constant, koff is the unbinding rate constant, t0 is the time at the start of the dissociation phase and Rmax and C are as described in Eq. (1). The binding affinity is calculated as, KD ¼

koff kon

ð4Þ

Note that kinetic fits are only applicable for data that exhibits enough curvature to be fit accurately. For compounds with fast association (kon > 107 M1 s1) and fast dissociation (koff > 0.1 s1), kinetic fitting is not appropriate [6, 12]. 4. In this experiment, the reagent line will be used for the 50% DMSO wash solution. Prepare 500 mL 50% DMSO in Milli-Q water. Insert reagent line into the bottle and ensure this line is properly primed. 5. For all wash steps, only the flow system and injection needles are cleaned. The wash solution does not flow over the sensor surface. 6. The immobilization contact time was optimized in a separate experiment to achieve the desired amount of protein on the surface. Ideally, the immobilization density should be as low as possible to achieve the best reference subtraction and mitigate re-binding events and mass transport limitation but large enough to ensure reliable signal to noise. The signal window is determined by the maximum binding response at saturation,

Protein Ligand Interactions by SPR

17

Rmax. A theoretical value of Rmax can be calculated for a 1:1 binding interaction on a fully active surface using the equation: Rmaxtheory ¼ Rimmobilized

MW analyte MW Ligand

ð5Þ

Rimmobilized is the immobilization density, MWanalyte is the molecular weight of the compound and MWligand is the molecular weight of the protein on the surface. Setting Rmax_theory between 20 and 50 RU typically gives a sufficient window. For this experiment, the molecular weight of mCRBN-TDB is 14.3 kDa and the average molecular weight of the compounds is 500 Da, thus, the targeted protein immobilization density is 1430 RUs for a 50 RU Rmax_theory. The average capture density achieved was 2400 RU. 7. In this experiment, all compounds are injected with the same contact and dissociation time. However, this can be made a variable and can be defined for each compound in the VARI ABLES AND POSITIONING tab. 8. For this screen, a regeneration step was not included, but can be added in the event of a slowly dissociating compound. Commonly, a high salt injection is used for regeneration, but other solutions can be used. Regeneration scouting is recommended to find the appropriate solution that removes compounds but does not perturb protein function. 9. The positive control is important for monitoring surface activity as a function of time. It is common to see the positive control binding signal decay throughout the course of the experiment as the surface activity decreases with time. It is good practice to run a titration of a tool compound at the beginning and end of the experiment to make sure that the binding affinity and kinetics are the same indicating that the decay in surface activity is tolerable (Rmax may be lower at the end of the experiment). It is also good practice to run a longevity study during assay development to determine how long the protein stays active on the surface. In this study of only 5 compounds, the surface remains active through the course of the experiment. 10. The negative control can be used to determine the noise in the system. Signals below 3*standard deviation of the negative control signal should be considered insignificant. Compounds that do not produce data above this cut-off throughout the dose response should be considered non-binders. 11. Solvent correction is used to correct for the volume exclusion effect when reference subtracting samples [12]. In this experiment, solvent correction was run at the beginning and end of the experiment since only five compounds were screened

18

Nichole O’Connell

(48 cycles of compound injections). The loop counter was set to every 400 cycles to trick the system into running only 2 solvent correction cycles. For larger experiments of one or more plates, it is recommended to run an additional solvent correction in the middle of the experiment (select DISTRIB UTE ONE CYCLE EVENLY). Four solvent correction samples were used that span 0.5% above and below the assay buffer DMSO concentration of 1%. Samples were prepared at 0.5%, 0.8%, 1.2%, and 1.5% DMSO. 12. Since only four channels are used in this method, the software will populate well positions in the upper half of the plate (rows A–H for a 384 well plate and rows A–D for the 96 well plate). However, solution must also be placed in the lower half of the plate because the instrument has eight fixed needles that will draw from these wells during the experiment. Having water or assay buffer in the unused wells prevents air from being injected into the system. 13. The software automatically calculates the volume of each solution required, accounting for dead volume. 14. Note that samples that are injected during a single cycle must reside on the same tray. For example, if a compound injection from Plate 1 is followed by a regeneration injection, the regeneration solution must reside on the same tray, either in Plate 1 or Plate 2. 15. The latest version of Control Software contains a Volume summary feature which defines how much solution is needed to fill all wells of a given sample. This is very useful for calculating the required volume needed to fill the support plate. It is wise to make slightly more solution than recommended so you have some dead volume in the source tube. 16. Use only the Biacore-approved plate foils as they do not have glue covering the wells, where the needle inserts. This is critical to avoid as the glue may contaminate the needle and give rise to signal artifacts. 17. Predefined methods perform standard tasks of data fitting, such as normalizing the curves to the baseline report point, performing double referencing through reference and blank subtraction, calculating solvent correction curves, and defining plots to QC the data. 18. The four solvent correction injections should produce rectangular bulk shifts that are negative for samples 1 and 2 and positive for samples 3 and 4. Blue lines in the response curve plot above the sensorgram indicate the bounds of all the data collected in the experiment. These lines should fall within the

Protein Ligand Interactions by SPR

19

bounds of the solvent correction curves. If data falls outside the bounds, those sensorgrams will be considered “outside the range of solvent correction.” One can extrapolate to apply solvent correction to outlying data, but it is not encouraged. 19. The default is to display the “corrected” sensorgrams from the Analysis step of the method. To look at the raw data for the active or reference channel, or to look at the reference subtracted, uncorrected data, open the SELECT SENSOR GRAMS to the left of the screen under SETTINGS and change the selection under SENSORGRAM TYPE. This panel will also allow you to look at data from other steps in the method, such as the immobilization step. 20. If there is significant drift in the baseline, steps must be taken to correct this by either equilibrating the system for a longer time or changing the immobilization method. In general, immobilizing through streptavidin capture of a biotinylated protein is a high affinity interaction that is stable throughout the screen. Typical baseline drift in Biacore systems is < 0.3 RU/min. 21. The ANALYTE STABILITY EARLY report point is taken in the dissociation phase, 5 s after the end of an injection (see Fig. 1). Upon analyte injection, signal on the reference flow cell should purely be due to bulk shift and should decay immediately after the injection is finished, leaving a value close to zero for this report point. Non-zero signal observed on the reference cell at this point indicates that the compound is sticking to the chip surface. In general, minor reference binding can be tolerated, but large reference binding (>5–10 RU) will cause artifacts in reference subtracted data. 22. Take care to understand the source of decay in the positive control signal over the course of experiment. In this case, binding to the active flow cell is consistent over the course of the experiment. The signal decay is due to increased binding to the reference channel over the course of the experiment. This artifactually manifests as a decay in the surface activity. It is unclear why Pomalidomide binding to the reference increases over time. 23. The software will automatically fit the data; however, it is prudent to go through each compound to ensure the fit is done using the highest quality data. 24. The kinetic rate constants kon and koff are equivalent to ka and kd, respectively.

20

Nichole O’Connell

Acknowledgements The author thanks A. J. H and H. L. for intellectual contributions and edits, R. S. for stellar Biacore support and C.M.Y for help rendering compound structures. References 1. Hanson SM, Georghiou G, Thakur MK, Miller WT, Rest JS, Chodera JD, Seeliger MA (2019) What makes a kinase promiscuous for inhibitors? Cell Chem Biol 26(3):390–399.e5. https://doi.org/10.1016/j.chembiol.2018. 11.005 2. Roy MJ, Winkler S, Hughes SJ, Whitworth C, Galant M, Farnaby W, Rumpel K, Ciulli A (2019) SPR-measured dissociation kinetics of PROTAC ternary complexes influence target degradation rate. ACS Chem Biol 14 (3):361–368. https://doi.org/10.1021/ acschembio.9b00092 3. Smith BE, Wang SL, Jaime-Figueroa S, Harbin A, Wang J, Hamman BD, Crews CM (2019) Differential PROTAC substrate specificity dictated by orientation of recruited E3 ligase. Nat Commun 10(1):1–13. https://doi. org/10.1038/s41467-018-08027-7 4. Zorba A, Nguyen C, Xu Y, Starr J, Borzilleri K, Smith J, Zhu H, Farley KA, Ding W, Schiemer J, Feng X, Chang JS, Uccello DP, Young JA, Garcia-Irrizary CN, Czabaniuk L, Schuff B, Oliver R, Montgomery J, Calabrese MF (2018) Delineating the role of cooperativity in the design of potent PROTACs for BTK. Proc Natl Acad Sci U S A 115(31): E7285–E7292. https://doi.org/10.1073/ pnas.1803662115 5. Markey F (2000) Principles of surface Plasmon resonance. In: Real-time analysis of biomolecular interactions. Springer Japan, Tokyo, pp 13–22. https://doi.org/10.1007/978-4431-66970-8_2 6. Schasfoort RBM (2017) Chapter 1. Introduction to surface plasmon resonance. In: Handbook of surface plasmon resonance. Royal Society of Chemistry, London, pp 1–26. https://doi.org/10.1039/978178801028300001 7. Chamberlain PP, Lopez-Girona A, Miller K, Carmel G, Pagarigan B, Chie-Leon B, Rychak E, Corral LG, Ren YJ, Wang M, Riley M, Delker SL, Ito T, Ando H, Mori T,

Hirano Y, Handa H, Hakoshima T, Daniel TO, Cathers BE (2014) Structure of the human Cereblon–DDB1–lenalidomide complex reveals basis for responsiveness to thalidomide analogs. Nat Struct Mol Biol 21(9):803–809. https://doi.org/10.1038/nsmb.2874 8. Mori T, Ito T, Liu S, Ando H, Sakamoto S, Yamaguchi Y, Tokunaga E, Shibata N, Handa H, Hakoshima T (2018) Structural basis of thalidomide enantiomer binding to cereblon. Sci Rep 8(1):1294. https://doi. org/10.1038/s41598-018-19202-7 9. Yamshon S, Ruan J (2019) IMiDs new and old. Curr Hematol Malig Rep 14(5):414–425. https://doi.org/10.1007/s11899-01900536-6 10. Winter GE, Buckley DL, Paulk J, Roberts JM, Souza A, Dhe-Paganon S, Bradner JE (2015) Phthalimide conjugation as a strategy for in vivo target protein degradation. Science 348(6241):1376–1381. https://doi.org/10. 1126/science.aab1433 11. Piya S, Bhattacharya S, Mu H, Lorenzi PL, McQueen T, Davis ER, Ruvolo V, Baran N, Qian Y, Crews C, Kantarjian HM, Andreeff M, Borthakur G (2016) BRD4 proteolysis targeting chimera (PROTAC) ARV-825, causes sustained degradation of BRD4 and modulation of chemokine receptors, cell adhesion and metabolic targets in leukemia resulting in profound anti-leukemic effects. Blood 128(22):748. https://doi.org/ 10.1182/blood.v128.22.748.748 12. Biacore Assay Handbook 29-0194-00 Edition AA, Cytiva Life Sciences. https://www. cytivalifesciences.co.jp/contact/pdf/ BiacoreAssayHandbook.pdf 13. Bondeson DP, Smith BE, Burslem GM, Buhimschi AD, Hines J, Jaime-Figueroa S, Wang J, Hamman BD, Ishchenko A, Crews CM (2018) Lessons in PROTAC design from selective degradation with a promiscuous warhead. Cell Che Biol 25(1):78–87.e5. https:// doi.org/10.1016/j.chembiol.2017.09.010

Chapter 2 High-Throughput Detection of Ligand-Protein Binding Using a SplitLuc Cellular Thermal Shift Assay Tino W. Sanchez, Ashley Owens, Natalia J. Martinez, Eric Wallgren, Anton Simeonov, and Mark J. Henderson Abstract The confirmation of a small molecule binding to a protein target can be challenging when switching from biochemical assays to physiologically relevant cellular models. The cellular thermal shift assay (CETSA) is an approach to validate ligand-protein binding in a cellular environment by examining a protein’s melting profile which can shift to a higher or lower temperature when bound by a small molecule. Traditional CETSA uses SDS-PAGE and Western blotting to quantify protein levels, a process that is both time consuming and low-throughput when screening multiple compounds and concentrations. Herein, we outline the reagents and methods to implement split Nano Luciferase (SplitLuc) CETSA, which is a reporter-based target engagement assay designed for high-throughput screening in 384- or 1536-well plate formats. Key words CETSA, Cellular Thermal Shift Assay, Thermal stability, Target engagement, Luciferase, Ligand binding, High-throughput screening, SplitLuc, qHTS CETSA

1

Introduction Drug discovery often relies on the confirmed interaction between small molecules and their proposed target protein. However, the failure of many lead compounds relates to poor physicochemical and pharmacokinetic properties that are associated with the native cellular environment [1]. Ligand-target engagement in a cellular environment can be confirmed using a cellular thermal shift assay (CETSA) that reports quantifiable changes in the thermal stability of a target protein when bound by a small molecule [2]. In a conventional CETSA, cells are incubated with a vehicle control or test compound (putative ligand) and heated at different temperatures, followed by Western blotting to examine thermal stability of

Supplementary Information The online version of this chapter (https://doi.org/10.1007/978-1-0716-16659_2) contains supplementary material, which is available to authorized users. Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

21

22

Tino W. Sanchez et al.

A 37

50

55

60

65

70

75

– + – + – + – + – + – + – +

Temp oC LDHAi anti-LDHA anti-SOD1

B Signal (relative to 37 oC)

150

100

50

Tagg DMSO 58.6 LDHAi 69.5

0 40

50

Temperature

60

70

(oC)

Fig. 1 Traditional CETSA using Western blotting. (a) HEK293T cells were treated for 1 h with 10 μM LDHA inhibitor (LDHAi) and soluble protein was detected by immunoblot. SOD1 was used as a thermally stable loading control. (b) Densitometry of immunoblots (mean, N ¼ 2 per condition)

the protein of interest [3]. The premise is that high affinity compounds that access and bind to the target protein in a cellular environment can alter protein folding thermodynamics resulting in a shift in thermal stability (Fig. 1). Although a standardized CETSA protocol using antibodies to the target protein can show ligand-target engagement, the Western blotting procedure is laborious and slow. Pre-clinical drug discovery efforts would greatly benefit from high-throughput CETSA screening platforms to identify lead molecules or evaluate structure-activity relationships between hundreds of analogues during lead optimization [4]. Towards this goal, we have developed a split Nano luciferase (SplitLuc) CETSA platform to directly assess target engagement in a cellular environment [5, 6], a method that can be utilized to rapidly interrogate the binding of thousands of compounds to a target protein (Fig. 2). SplitLuc CETSA employs a transient expression of the target of interest (TOI) fused to a 15 amino acid Nano luciferase fragment, referred to as 86b (Gly-Ser-HiBiT-Gly-Ser). Cells expressing the TOI are incubated with small molecule ligands, heated, lysed, and analyzed for

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

23

A

LDHA

86b

B Soluble

N

N

+

+ 11S (LgBiT)

N OH O

Light

Complementation

C

Aggregate

Luminescence (relative to 37 oC)

100

TEMPERATURE

Tagg oC

50

DMSO 62.6 LDHAi 72.0 0 40

60

80

Temperature (oC)

E

TOI stabilization (% vs unheated)

D

NO HEAT

HEAT

100

50

0 -9

-8

-7

-6

-5

Log [Compound] (M) COMPOUNDS COMPOUNDS

Fig. 2 (a) Schematic depicting LDHA-SplitLuc CETSA approach. 86b (HiBiT colored in red, with Gly-Ser amino acids on both ends) is appended to TOI and luminescence is detected after the addition of 11S and furimazine. (b) At high temperatures, complementation does not occur as protein unfolds and the peptide tag becomes buried within aggregates (composed of many cellular proteins). Centrifugation is not required. (c) Example of temperature-response profile for SplitLuc CETSA. LDHA-86b transfected cells were treated with DMSO or 10 μM LDHA inhibitor for 1 h. Cells were heated for 3.5 min to the indicated temperatures and soluble protein was assessed using the SplitLuc CETSA approach (mean, N ¼ 4). (d) Example of an isothermal 1536-well SplitLuc CETSA assay. Unheated (top) and heated (bottom) plates are shown. Cells were treated with small molecule inhibitors in a dose-response for 1 h before performing the SplitLuc CETSA. (e) Example of dosedependent stabilization for TOI, as depicted in panel d

24

Tino W. Sanchez et al.

luminescence using a complementary Nano luciferase enzyme fragment to detect binding-induced thermal shifts. Newer trends in small molecule drug discovery are moving beyond traditional targeting of protein active sites to also pursue allosteric binding sites or degradation-targeting strategies [7– 9]. The high-throughput CETSA protocol outlined below provides a unique and widely applicable approach to identify small molecule binders that have a high affinity for their target protein regardless of whether the binding event alters enzymatic activity. Small molecules identified in SplitLuc CETSA could provide starting points for developing proteolysis targeting chimeras (PROTACs) that degrade a protein target and contribute to unlocking a new set of “druggable” features of proteins [10]. We expect the growing field of protein degradation strategies, which also include lysosomal (LYTAC), autophagy (AUTAC) and endosomal (ENDTAC) targeting chimeras, will greatly benefit from high-throughput CETSA platforms, and these technologies will expand the druggable genome and offer new opportunities for drug discovery [11–13].

2

Materials Store materials at room temperature unless otherwise specified.

2.1 SDS PAGE and Western Blotting

1. 4–12% Bis-Tris polyacrylamide gel: store at 4  C. 2. 20 SDS-PAGE buffer: 50 mM MOPS, 50 mM Tris Base, 0.1% SDS, 1 mM EDTA, pH 7.7. 3. Electrophoresis apparatus. 4. 4 loading buffer: 240 mM Tris–HCl, pH 6.8, 8% SDS, 40% Glycerol, 5% beta-mercaptoethanol, 0.04% bromophenol blue. 5. Molecular weight protein ladder: store at 20  C. 6. PVDF membrane (e.g., iBlot stack). 7. Transfer apparatus (e.g., iBlot2). 8. Tris Buffered Saline (TBS): 50 mM Tris–Cl, pH 7.5, 150 mM NaCl. 9. Wash buffer: TBS with 0.05% tween-20 (TBS-T). 10. Blocking Buffer: 5% milk in TBS-T, store at 4  C. 11. Monoclonal primary antibody: store according to manufacturer’s instructions (e.g., 20  C). 12. Secondary antibody: store according to manufacturer’s instructions (e.g., at 20  C). 13. Enhanced Chemiluminescence (ECL). 14. Western blot imager (e.g., ChemiDoc, Typhoon, LiCor).

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

25

N-TERM ACCEPTOR kozak

start 86b (Met-Gly-Ser-HiBiT)

BamHI (Gly-Ser)

ACCCAAGCTGGCTAGCCACCatgGGCAGCGTGAGTGGCTGGCGACTGTTCAAGAAGATCAGCGGATCCaactaacaatagcgttatcGAATTCTGCAGATAT

C-TERM ACCEPTOR NheI site

BamHI (Gly-Ser)

remainder 86b (HiBiT-Gly-Ser-STOP)

EcoRI

ACCCAAGCTGGCTAGCaactaacGGATCCGTGAGTGGCTGGCGACTGTTCAAGAAGATCAGCGGCAGCTAAGGCGCGCCGAATTCTGCAGATAT

pcDNA3.1 (+) digested with NheI and EcoRI

Fig. 3 Schematic for SplitLuc Acceptor plasmid design. This plasmid facilitates cloning of TOIs, with N- or C-terminal reporter tag 2.2

Compounds

1. DMSO. 2. Small molecules: compounds dissolved in DMSO at 10 mM, store stock solutions at 20  C.

2.3 Constructing the SplitLuc CETSA Acceptor Plasmids

1. pcDNA3.1(+) acceptor plasmid. 2. 86b (Gly-Ser-HiBiT-Gly-Ser) insert sequence to create N-terminal acceptor plasmid, with overlap to NheI/EcoRI linearized pcDNA3.1 (see Fig. 3): ACCCAAGCTGGCTAGC CACCATGGGCAGCGTGAGTGGCTGGCGACTGTTCAA GAAGATCAGCGGATCCAACTAACAATAGCGTTATCGA ATTCTGCAGATAT. Use complementary oligonucleotides to create double-stranded DNA. 3. 86b (Gly-Ser-HiBiT-Gly-Ser) insert sequence to create C-terminal acceptor plasmid, with overlap to NheI/EcoRI linearized pcDNA3.1 (see Fig. 3): ACCCAAGCTGGCTAG CAACTAACGGATCCGTGAGTGGCTGGCGACTGTTCA AGAAGATCAGCGGCAGCTAAGGCGCGCCGAATTCTGCAGATAT. Use complementary oligonucleotides to create double-stranded DNA. 4. NheI: store at 20  C. 5. EcoRI: store at 20  C.

26

Tino W. Sanchez et al.

6. 1% agarose gel with SYBR-Safe stain (or other DNA stain). 7. Agarose gel electrophoresis apparatus. 8. Gel extraction spin columns (e.g., Qiagen, Machery-Nagel). 9. 10 Annealing buffer for oligonucleotides: 10 mM Tris HCl, pH 8, 500 mM NaCl, 10 mM EDTA, store at 20  C. 10. In-Fusion HD EcoDry Cloning reaction kit: once package is opened, store in desiccator. 11. Stellar competent cells: store at 80  C and do not allow to thaw prior to use. 12. SOC medium: 0.5% yeast extract, 2% tryptone, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4. Autoclave the solution then add 20 mM glucose when cool. Filter the solution to sterilize. 13. Transformation enhancing buffer: 5 mM Tris-HCl, pH 8.5. 14. LB agar plates: Autoclave LB agar (1% tryptone, 0.5% yeast extract, 1% sodium chloride, 1.5% agar), once cool, add 100 μg/mL carbenicillin, pour plates, store at 4  C. 15. Round bottom culture tubes with loose fitting caps. 16. LB broth: 1% tryptone, 0.5% yeast extract, 1% sodium chloride. 17. Sterile spreaders or loops. 18. Miniprep kit. 19. Spectrophotometer for DNA quantification (e.g., NanoDrop). 20. pcDNA3.1 forward sequencing primer: GGCTAACTAGA GAACCCACTG. 21. pcDNA3.1 reverse sequencing primer: GGCAACTAGAAGG CACAGTC. 2.4 Cloning TOI into the SplitLuc CETSA Plasmid

1. BamHI: store at 20  C. 2. Linearized SplitLuc CETSA acceptor plasmid. 3. PCR primers for TOI (with N-terminal 86b tag). Forward: 50 - GAAGATCAGCGGATCC [sequence of TOI, starting with ATG]-30 . Reverse: 50 - ATATCTGCAGAATTC [reverse complement to TOI end, include STOP codon]-30 . 4. PCR primers for TOI (with C-terminal 86b tag). Forward: 50 -ACCCAAGCTGGCTAGCACC [sequence TOI, starting with ATG]-30 . Reverse: 50 -AGCCACTCACGGATCC [reverse complement to TOI end, NO STOP codon]-30 . 5. High Fidelity Taq polymerase. 6. Sequencing primers for TOI (internal to gene for large genes).

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

2.5 Cell Culture and Transfection

27

1. HEK293T cells (ATCC #CRL3216). 2. Culture medium: DMEM, 4.5 g/L glucose, 6 mM L-glutamine, 1 mM sodium pyruvate. 10% fetal bovine serum (FBS), 100 units/mL penicillin, 100 μg/mL streptomycin, store at 4  C. 3. 0.25% Trypsin/EDTA: store at 4  C. 4. Tissue culture-treated flasks and 6-well plates. 5. Tissue culture incubator: 37  C, 95% humidity, 5% CO2. 6. Transfection culture medium: phenol red-free DMEM, 4.5 g/L glucose, 6 mM L-glutamine, 10% FBS, store at 4  C. 7. OptiMEM. 8. Lipofectamine 2000: store at 4  C. 9. 0.3 μg of plasmid DNA per cm2 of tissue culture container, store at 4  C. 10. 1 Phosphate Buffer Saline without magnesium chloride and calcium chloride (PBS). 11. 0.25% phenol red-free trypsin/EDTA: store at 4  C.

2.6 CETSA and SplitLuc CETSA

1. Gradient capable 96-well or 384-well thermocycler. 2. PCR tubes. 3. Refrigerated centrifuge with PCR tube rotor. 4. CETSA assay buffer: phenol red-free DMEM, 4.5 g/L glucose, 6 mM L-glutamine, store at 4  C. 5. 6 lysis buffer: NP-40 diluted in deionized water to 6%, 6 protease inhibitor. 6. Furimazine, store stock at 20  C. 7. Recombinant 11S Nano luciferase fragment: store at 80  C. 8. 2 complementation reagent: 2 furimazine, 200 nM recombinant 11S Nano luciferase fragment diluted in CETSA assay buffer. 9. White cyclic olefin polymer 1536-well plates. 10. Plate reader equipped for luminescence detection (e.g., ViewLux, Envision, PHERAstar).

2.7 Metal Blocks and Heating Platform for Conductive Heat Transfer (Fig. S1)

1. Copper heat block (110 purity level copper sheet and rod) machined to fit the bottom of a standard 1536-well assay plate. 2. Adhesive-back polyamide film sheet heater. 3. Adhesive-back silicone foam bumpers. 4. Stainless steel threaded rod and threaded knobs. 5. High temperature silicone foam sheet with smooth texture cut to the dimensions of heat block.

28

Tino W. Sanchez et al.

6. Copper or aluminum clamping plate cover with same dimensions as heat block. 7. Thermocouple probe. 8. Programmable temperature controller and medium current relay. 9. Polycarbonate washdown enclosure with knockouts. 10. Adapter cord and power cord. 2.8

Counterscreens

1. 86b peptide ( GSVSGWRLFKKISGS ) dissolved in ultrapure water to 1 mM. Store at 80  C. 2. HEK293T cell lysate.

3

Methods Carry out all procedures at room temperature unless otherwise specified.

3.1 Traditional CETSA Using Western Blot Electrophoresis to Assess the Melting Profile of the TOI

1. Seed five million HEK293T cells in a T75 flask using prewarmed tissue culture medium and grow for 48 h at 37  C, 5% CO2, 95% RH. Harvest by aspirating the medium, rinse with 5 mL of 1 PBS, and incubate with 1–3 mL of 0.25% Trypsin/EDTA for 5 min in a 37  C tissue culture incubator. Quench trypsinized cells with 2–6 mL of warm tissue culture medium and spin for 5 mins at 500  g. Remove supernatant and resuspend the pellet in DMEM (no serum) final concentration of 10 million cells/mL (see Note 1). 2. Add 1–3 μL of DMSO to 1 mL of cell suspension and incubate at 37  C for 1 h. For TOIs, where a positive control compound is available, add 1–3 μL of 10 mM compound to 1 mL the cell suspension (10–30 μM final concentration) and incubate at 37  C for 1 h (see Note 2). 3. Aliquot 50 μL of cells into 12 (DMSO vehicle) or 24 (DMSO vehicle and positive control dissolved in DMSO) PCR tubes to match the different heating temperatures. Mount the PCR tubes in a gradient capable thermocycler and heat for 3 min— in pairs if incubated with compound—at temperatures ranging from 37 to 81  C in 4  C increments (see Notes 3–5). 4. After heating, allow the PCR tubes to cool to room temperature for 3 min. Add lysis buffer to each tube (6% NP40 and 6 protease inhibitor cocktail, to a final concentration of 1% and 1, respectively). Mix well by pipetting up and down and incubate for 30 min. Spin samples for 10 min at 15,000  g in a refrigerated centrifuge set at 4  C. Place the centrifuged

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

29

tubes on ice and transfer the clarified lysate into separate tubes for loading onto a PAGE gel. 5. To 15 μL of the clarified lysate, add 5 μL of 4 loading buffer and mix by pipetting up and down. Load a molecular weight protein ladder and 20 μL of the samples onto a 15-well 4–12% Bis-Tris PAGE gel. Fill the electrophoresis chamber with SDS-PAGE running buffer and run until proteins are separated (see Note 6). 6. Transfer the gel to a PVDF or nitrocellulose membrane (e.g., for 7 min at 25 V using an iBlot dry transfer system). Next, block the membrane with blocking buffer for 1 h. The membrane is then incubated with an antibody specific to the TOI according to manufacturer’s recommendations (see Note 7). 7. The next day, the membrane is washed three times by rocking with TBS-T for 10 min. After washing, incubate the membrane with a secondary antibody for 1 h on a rocking platform. The membrane is subsequently washed three more times by rocking with TBS-T for 10 min each wash (see Note 8). 8. After washing, the TOI expression is visualized using chemiluminescence or fluorescence readers, depending on the secondary antibody selected. The resulting image is analyzed for a thermal melt profile. Plot the temperature response profile for DMSO and control-treated samples, combining technical replicates (mean  SD), and calculate the Tagg, which is defined as the temperature at which a reduction of 50% signal (soluble protein) is observed. 3.2 Creating a SplitLuc “Acceptor” Plasmid for Rapid Cloning of SplitLuc Fusion Proteins

1. “Acceptor” plasmids can be constructed to facilitate the cloning of multiple and different TOIs with the SplitLuc 86b fragment appended to the N- or C-terminus (see Fig. 3, Note 9). 2. Anneal two complementary oligonucleotides to construct a double stranded 86b DNA fragment for insertion into the backbone pcDNA3.1 plasmid (N-terminal or C-terminal versions). Combine 5 μL of 100 μM forward oligonucleotide (Fig. 3), 5 μL of 100 μM reverse complement oligonucleotide, 5 μL of 10 annealing buffer and 35 μL of sterile water. 3. Bring 400 mL of water to a boil in a beaker on a hot plate. Place the tube of annealing oligonucleotide mix in the boiling water and incubate for 2 min, then turn off the hot plate and allow everything to equilibrate to room temperature. Move tubes to an ice bucket and store at 20  C until ready for use. 4. Cut 2 μg of pcDNA3.1(+) plasmid by incubating 10 U of NheI and 10 U of EcoRI in a 50 μL reaction volume for 1 h at 37  C. Load the digested plasmid product, across multiple wells, and an uncut plasmid control onto a 1% agarose gel and separate by

30

Tino W. Sanchez et al.

electrophoresis. Cut out the band of the large linearized plasmid (5.3 kb) and perform a gel extraction using a spin column, according to the manufacturer’s protocol (see Note 10). 5. Combine 30 fmol of the 86b-containing DNA fragments from step 3 (N-terminal or C-terminal) and 15 fmol of the linearized backbone plasmid from step 4 in a total volume of 10 μL in water. Also prepare a control reaction with 15 fmol of linearized backbone alone in a total volume of 10 μL in water. Add the mixtures to separate In-Fusion EcoDry lyophilized enzyme pellets for ligation. Incubate the mixture at 37  C for 15 min and then at 50  C for 15 min. Place on ice. 6. Add 35 μL of transformation enhancing buffer to the sample and pipet up and down on ice. Next, transform 50 μL of Stellar Competent bacterial cells by adding 1–4 μL of the mixture and incubate for 30 min on ice (see Note 11). 7. Heat shock the sample in a 1.5 mL microcentrifuge tube for exactly 60 s at 42  C in a water bath and then place the tube on ice for 2 min. Add 450 μL of SOC medium and shake the bacterial sample at 300 rpm for 1 h at 37  C in a mini shaker. 8. During the 1-h incubation, prewarm LB agar plates that have 100 μg/mL carbenicillin antibiotic in a 37  C incubator. Spread 100 μL of bacteria on a pre-warmed LB agar plate with carbenicillin using a sterile cell spreader. Incubate for 1 min to allow agar to soak up bacteria and incubate plate upside-down in a 37  C incubator overnight (see Note 12). 9. On the next day, ensure the transformed negative control has substantially fewer colonies compared to the In-Fused 86b-containing backbone. Select single colonies with a sterile pipet tip and inoculate 2 mL of LB broth with carbenicillin at 100 μg/mL (final concentration) in a mini culture tube. Grow the mini cultures overnight at 200 rpm in a 37  C incubator (see Note 13). 10. Perform a mini prep according to the manufacturer’s protocol to isolate the plasmid DNA. Calculate the concentration of the isolated DNA plasmid using a spectrophotometer. Sequence the sample using pcDNA3.1 forward and reverse sequencing primers to identify the correct acceptor plasmid clones. 11. Prepare a larger scale DNA preparation (e.g., midi or maxi) of the confirmed acceptor clones. 3.3 Cloning the TOI into a SplitLuc “Acceptor” Plasmid

1. The TOI gene to be inserted into the SplitLuc “acceptor” plasmid can be produced by DNA synthesis (e.g., gene strand) or PCR amplification (see Fig. 4, Note 14). 2. To linearize the N-terminal SplitLuc acceptor plasmid (from Subheading 3.2, step 11), add 10 U of BamHI and 10 U of

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

31

A EcoRI

BamHI (Gly-Ser)

GAAGATCAGCGGATCC--TARGET OF INTEREST (INCLUDE STOP CODON) -- GAATTCTGCAGATAT

PCR amplicon or synthetic gene fragment

pcDNA3.1 (+) N-term 86b TOI acceptor

linear fragment BamHI and EcoRI digest

B NheI

BamHI

ACCCAAGCTGGCTAG--TARGET OF INTEREST (INCLUDE CACCATG, kozak + START) -- GGATCCGTGAGTGGCT

PCR amplicon or synthetic gene fragment

pcDNA3.1 (+) C-term 86b TOI acceptor linear fragment NheI and BamHI digest

Fig. 4 Schematic for cloning TOIs into SplitLuc acceptor plasmid. (a) N-terminal design. (b) C-terminal design

EcoRI to 2 μg of plasmid in 50 μL of total volume and incubate for 1 h at 37  C. To linearize the C-terminal SplitLuc acceptor plasmid (from Subheading 3.2, step 11), add 10 U of NheI and 10 U of BamHI to 2 μg of plasmid and incubate for 1 h at 37  C (Fig. 3).

32

Tino W. Sanchez et al.

3. Load the digested plasmid products and an uncut plasmid control onto a 1% agarose gel and separate by electrophoresis. Cut out the band of the large linearized plasmid and perform a gel extraction and DNA purification according to the manufacturer’s protocol (see Note 10). 4. Mix 30 fmol of the target gene insert (gene strand or PCR amplicon) and 15 fmol of the linearized SplitLuc backbone plasmid from steps 1 and 3 in a total volume of 10 μL in water. Add the 10 μL to an EcoDry pellet of lyophilized enzyme for ligation. Incubate at 37  C for 15 min and then at 50  C for 15 min. Place on ice (see Note 15). 5. Add 35 μL of transformation enhancing buffer to the sample and pipet up and down on ice. Next, perform a bacterial transformation, colony selection and mini prep of the new TOI-SplitLuc DNA plasmid as described above in Subheading 3.2, steps 4–10. 6. Calculate the concentration of the isolated DNA using a spectrophotometer. Sequence sample using forward and reverse sequencing primers to identify the correct clones (see Note 16). 7. Perform a larger scale endotoxin-free DNA purification (e.g., midi or maxi) of sequence verified clones. 3.4 Transfection of HEK293T Cells for Validation of TOI-SplitLuc Expression

1. After confirming the correct sequence of the TOI and SplitLuc tag, transfect HEK293T cells with the newly constructed plasmid. For transfections done in a 6-well plate format, pre-incubate 3 μg of plasmid DNA in 150 μL of OptiMEM media in a 1.5 mL microcentrifuge tube and 6 μL of Lipofectamine 2000 with 150 μL of OptiMEM in a separate 1.5 mL tube for 5 min. Combine the two tubes and incubate for 20 min (see Note 17). 2. Seed 1 mL of 1–2 million HEK293T cells per well in a 6-well plate with transfection culture media. Carefully add ~300 μL of Lipofectamine-DNA complex in OptiMEM from step 1 to each well. Incubate the 6-well plate in a 37  C incubator for 24–48 h (see Note 18). 3. After transfection, remove supernatant and add 1 mL of 0.25% phenol red-free trypsin per well and incubate for 5 min in the 37  C incubator. Quench the trypsin in each well by adding 2 mL of transfection culture media and spin at 500  g for 5 min. Discard the supernatant and homogeneously resuspend the cell pellet to 1 million cells/mL in CETSA assay buffer (see Notes 19 and 20). 4. Transfer 30 μL of the resuspended cell sample to a PCR tube and add 6 μL of 6 lysis buffer. Incubate for 30 min.

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

33

5. Samples can be transferred to a new plate before performing a luciferase complementation assay. For example, aliquot 10 μL of each sample into a 384-well low volume white plate in triplicate. Add 10 μL of the complementation reagent (100 nM 11S luciferase fragment and 1 furimazine final concentration, diluted in CETSA assay buffer). Spin the plate at 300  g for 5 s and read on a plate reader equipped with luminescence detection (see Notes 21 and 22). 3.5 Low-Throughput SplitLuc CETSA (PCR Tube or 96-Well Plate)

1. Into a 1.5 mL microcentrifuge tube, incubate 1 mL of resuspended cells from Subheading 3.4, step 2 with a control compound ligand or matching vehicle control for 1 h at 37  C (see Note 4). 2. After 1 h, aliquot 30 μL of transfected cells from each 1.5 mL tube into PCR tubes or plates compatible with a 96 well block. Mount the PCR tubes from each condition (compound or DMSO control) onto a PCR thermocycler and heat tube separately for 3 min using a linear gradient from 37–88  C in 3  C increments (see Notes 5 and 23). 3. After heating, let the PCR tubes cool to room temperature for 3 min. Add 6 μL of 6 lysis buffer, mix well by pipetting up and down and incubate for 30 min. If heating is performed in a white PCR plate, add 36 μL of 2 complementation reagent directly to plate. Alternatively, transfer equivalent volume of samples from PCR tubes to white 96- or 384- well plates. For example, transfer 10 μL of the supernatant into a white 384-well plate (technical replicates of 10 μL each from each sample can be added). 4. Spin the plate gently at 300  g for 5 s. Add 10 μL of 2 complementation reagent. Spin the plate again at 300  g for 5 s. Read on a plate reader equipped with luminescence detection (see Note 21). 5. Plot the temperature response profile of DMSO and controltreated samples, combining technical replicates (mean  SD calculated from independently heated samples), and calculate Tagg. The temperature at which 70–80% decrease of luminescent signal is observed is also calculated (see Note 24).

3.6 High-Throughput SplitLuc CETSA (384-Well Format)

1. Following the same protocol in Subheading 3.4, transfect 10–20 million HEK293T cells in 10 mL of transfection culture media in a T75 flask using 45 μL lipofectamine in 500 μL of OptiMEM complexed with 22.5 μg of plasmid DNA (SplitLuc TOI) in 500 μL of OptiMEM. Scale accordingly if more cells are needed. 2. After 24–48 h, aliquot 10 μL of transfected cells resuspended in CETSA assay buffer into each well of a white 384-well PCR

34

Tino W. Sanchez et al.

plate (see Notes 25 and 26). Then add DMSO controls or desired compound concentrations and incubate for 1 h at 37  C. 3. Seal plate with optical film and heat in a PCR block for 3 min at a temperature that corresponds to 70–80% decrease of signal observed in Subheading 3.5 (see Notes 23, 27–30). 4. Let the plate cool to room temperature for 10 min. Add 2 μL of 6 lysis buffer to the 384-well plate. Spin the plate at 300  g for 5 s and incubate for 30 min. Next, add 12 μL/well of 2 complementation reagent to the 384-well plate. 5. Spin the plate at 300  g for 5 s and read on a plate reader equipped for luminescence (see Note 23). 3.7 High-Throughput SplitLuc CETSA (1536-Well Format)

1. Assay can be performed using adherent or suspension cells. For suspension, aliquot 5 μL/well of transfected cells (2  105 to 1  106 cells per mL) resuspended in CETSA assay buffer from protocol in Subheading 3.4 into a white, 1536-well plate using a Multidrop Combi Reagent dispenser. Immediately proceed with step 2. For adherent cells, add 5 μL of cells (1000–2000 cells) into a white, 1536-well plate and incubate overnight. 2. Optimize the melting conditions for 1536-well format (see Note 31). For optimization, cells can be scattered across the plate (e.g., plate cells in one out of every four columns) to confirm even melting across the plate, while saving reagents. Optimize temperature by incubating the 1536-well plates on a copper block set to temperatures that bracket the previously observed melting Tagg of the TOI (see Note 27). Incubation time (e.g., 6, 9, or 12 min) should also be optimized for each TOI. When available, include a positive control compound in the optimization experiments to track thermal stabilization under the different conditions and guide selection of assay conditions for subsequent experiments. Target a temperature and time that corresponds to 50–75% melt and decrease of luminescent signal. Calculate Z0 factors using heated and unheated samples (or compound treated if available). Select conditions that provide the most robust assay performance. 3. Before heating plates, seal with optical film or foil seal, place plate on a copper heat block and immediately apply even top pressure. Alternatively, float the plate in a water bath set to the desired temperature (see Notes 27–32). 4. Let the plate cool to room temperature for 10 min. Add 1 μL/ well of 6 lysis buffer to the 1536-well plate using a Multidrop Combi dispenser. Spin the plate at 300  g for 5 s and incubate for 30 min at room temperature. Next add 3 μL of complementation reagent (300 nM 11S and 3 furimazine, for a final concentration of 100 nM and 1, respectively).

High-Throughput SplitLuc CETSA for Ligand-Protein Binding

35

5. Spin the plate at 300  g for 5 s and read on a plate reader equipped for luminescence (see Note 21). 6. High-throughput CETSA experiments with hundreds or thousands of compounds can be performed as described in Subheading 3.7, steps 1–5, by adding compounds for 1–2 h before heating samples. 3.8 Counterscreens, Data Analysis, and Confirmation

1. For screening, include an unheated (37  C) control plate treated with compounds (under identical conditions used for the heated plate) to identify assay artifacts, including compounds that affect SplitLuc activity or complementation (see Note 29). 2. A decrease/increase in signal observed for samples in the unheated plate, however, should be interpreted with caution, as small molecule induced stabilization/destabilization of protein under physiologic temperature (and corresponding increase/decrease in luminescent signal) is a consequence of bona fide target engagement in some circumstances. 3. For compounds that show activity in the unheated plate, potential spurious effects on SplitLuc can be confirmed by testing SplitLuc components independent of the TOI. To perform this assay, first create HEK293T lysate (untransfected) using cell density that matches the high-throughput CETSA conditions. Then, spike in 86b peptide at 200 nM final concentration, add compounds at the desired concentration and incubate for 30 min. Next, add complementation reagent and measure luminescence. Activity in this assay indicates the compound is altering the reporter components, rather than the TOI. 4. To analyze the high-throughput CETSA data, first normalize every well from the heated plate using the corresponding well from the unheated plate (Value[heated]/Value[unheated]). The normalized data for the DMSO control wells will represent the percent melting that was achieved in the experiment, and this should match expectations before proceeding with further analysis. Calculate the mean and standard deviation for the DMSO samples. Compare the treated samples with DMSO to identify compound-induced stabilization or destabilization. Compounds that pass a threshold (mean melting for vehicle  3SD) are selected for follow-up confirmation (see Note 29). 5. Cherry pick compounds and re-test (replicates or doseresponse), following steps outlined above, to confirm TOI stabilization.

36

4

Tino W. Sanchez et al.

Notes 1. HEK293T cells are used as the default cell line because of their ease of growth and high transfection efficiency. Cells are used between passage 10 and 20, for consistency in transfection and cell growth. Other transfectable cell lines can be substituted. For some targets, melt profiles and target engagement will not be identical between cell lines (post-translational modification, protein-protein interactions, etc.). Cell lines that are relevant to the biology under investigation should be considered. 2. For TOIs where a positive control compound is available, use a concentration where maximal activity was observed in a cellular environment. If only biochemical data is available, then begin with a concentration around 30 μM if solubility and cytotoxicity are not an issue. 3. The rationale for making a SplitLuc construct is to show target engagement and thermal stabilization. Thus, before constructing a SplitLuc plasmid, its ideal to first identify the melting profile of the endogenous target protein and confirm compound-induced thermal shift with a control compound known to engage the target. Compound binding to a target does not always translate to thermal shift, and therefore, all CETSA formats are prone to false negatives. 4. When incubating compounds and DMSO vehicle control with cells, use matched concentrations of 20-fold binding affinity at the E3 ligase VHL (Kd 3 μM) compared to MZ1 [29] retained potent and selective degradation of Brd4 at concentrations well below the binary binding affinity (DC50 10 nM), as a result of the large cooperativity of the ternary complex [30]. More recently, Han et al. reported similar findings that PROTACs made of weak-affinity VHL ligands can still work as potent degraders [31]. Together these amongst other studies [32] have demonstrated feasibility of achieving potent degraders made of weak-affinity ligands. As PROTACs induce neo protein–protein interactions between the target protein and the E3 ligase, they can potentially target different surfaces of the partner proteins, and consequently cooperativity and stability of ternary complexes have been found to differ greatly between different ubiquitin ligases and different POIs [27, 33]. This allows for enhanced target selectivity but also makes the rational design of PROTACs difficult at the present stage [5]. Addressing these unknowns is best accomplished with robust in vitro high-

118

Tanja Bartoschik et al.

throughput screening approaches for both binary interactions and ternary complex formation. Here, we describe an experimental protocol to characterize the binary and ternary interactions of the MZ1 PROTAC with E3 ubiquitin ligase and a set of different BET bromodomain constructs using MST [29, 33, 34]. The entire experimental procedure, starting with assay development, followed by the MST measurement, and ending with data interpretation, is described in detail. The results show that MST is a powerful method for PROTAC development and can be used to rank involved molecules according to their cooperativity. Although the measurements shown in this chapter are based on a small set of POIs, MST can be applied for high-throughput screening projects by integrating the instruments into any liquid handling system [22].

2

Materials Prepare all solutions using ultrapure deionized water (dH2O) and analytical grade reagents. Prepare and store all buffers and all reagents at room temperature, unless stated otherwise. Keep proteins on ice during the experiments.

2.1 Protein Preparation

1. VHL (UniProt accession number: P40337).

2.1.1 Proteins and Constructs

3. ElonginB (UniProt accession number: Q15370).

2. ElonginC (UniProt accession number: Q15369). 4. Brd2BD2 (344-455). 5. Brd3BD1 (24-146). 6. Brd4BD1 (44-178). 7. Brd4BD2 (333-460).

2.1.2 Protein Production and Purification

1. Escherichia coli BL21(DE3). 2. Isopropyl β-D-1-thiogalactopyranoside (IPTG). 3. Pressure cell homogenizer (Stansted Fluid Power). 4. HisTrap FF affinity column (GE Healthcare). 5. Imidazole. 6. TEV protease. 7. MonoQ and Superdex-75 columns (GE Healthcare). 8. Nickel Sepharose 6 fast flow beads (GE Healthcare). 9. Sepharose 6 fast flow beads (GE Healthcare). ¨ kta FPLC purification systems (GE Healthcare). 10. A 11. Glass econo-columns (Bio-Rad).

MST and TRIC to Study PROTACs in Drug Discovery

2.2

Protein Labeling

119

1. NHS labeling buffer (NanoTemper Technologies GmbH, Germany). 130 mM NaHCO3 pH 8.2, 50 mM NaCl. 2. Monolith Protein Labeling Kit RED-NHS 2nd Generation (Amine Reactive) (NanoTemper Technologies GmbH, Germany). 3. DMSO (100%). 4. Vortexer. 5. Assay buffer (NanoTemper Technologies GmbH, Germany). 50 mM Tris–HCl pH 7.8, 150 mM NaCl, 10 mM MgCl2, 2 mM GSH, 0.05% Tween-20. 6. NanoDrop. 7. Benchtop centrifuge.

2.3 Affinity Measurements with MST

1. 384-well microtiter plate 2. BET PROTAC MZ1 (obtained from opnMe) stock solution: 5.1 mg suspended in 1 mL DMSO to yield a 51 mM suspension. 3. Monolith NT.115 Premium Capillaries (NanoTemper Technologies GmbH, Germany). 4. Monolith NT.115Pico (NanoTemper Technologies GmbH, Germany). 5. MO.Control Software (NanoTemper Technologies GmbH, Germany). 6. MO.Affinity Analysis Software (NanoTemper Technologies GmbH, Germany).

2.4

SD-Test

1. SD-mix (NanoTemper Technologies GmbH, Germany). 4% SDS, 40 mM DTT. 2. Nonbinding microcentrifuge tubes. 3. Heating block.

3

Methods Carry out all procedures at room temperature unless otherwise specified.

3.1 Protein Production and Purification

Protein production and purification was done by Dr. Klaus Rumpel and Dr. Andreas Zoephel at Boehringer Ingelheim, according to Morgan S Gadd et al. [27]. Briefly, wild-type and mutant versions of human proteins VHL (UniProt accession number: P40337), ElonginC (Q15369), ElonginB (Q15370), bromodomains (BDs) of BET proteins Brd2 (P25440), Brd3 (Q15059), and Brd4 (O60885)

120

Tanja Bartoschik et al.

Table 1 Mutant constructs and storage buffers Protein

Storage buffer

Ubiquitin ligase (VCB)

20 mM HEPES 7.5, 100 mM NaCl, 1 mM TCEP

BD2

Brd2

(344–455)

Brd3BD1 (24–146)

10 mM HEPES 7.5, 500 mM NaCl, 5% Glycerol 25 mM HEPES 7.5, 150 mM NaCl, 10 mM DTT

BD1

(44–178)

10 mM HEPES 7.5, 100 mM NaCl, 1 mM DTT

BD2

(333–460)

10 mM HEPES 7.5, 100 mM NaCl, 10 mM DTT

Brd4 Brd4

were used for all protein expression. For expression of VCB, N-terminally His6-tagged VHL (54–213), ElonginC (17–112) and ElonginB (1–104) were co-expressed in Escherichia coli BL21(DE3) at 24
C for 16 h using 3 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). E. coli cells were lysed using a pressure cell homogenizer (Stansted Fluid Power) and lysate clarified by centrifugation. His6-tagged VCB was purified on a HisTrap FF affinity column (GE Healthcare) by elution with an imidazole gradient. The His6 tag was removed using TEV protease and the untagged complex dialyzed into low-concentration imidazole buffer. VCB was then flowed through the HisTrap FF column a second time, allowing impurities to bind, as the complex eluted without binding. VCB was then additionally purified by anion exchange and size-exclusion chromatography using MonoQ and Superdex-75 columns (GE Healthcare), respectively. The final purified complex was stored in 20 mM HEPES 7.5, 100 mM NaCl, 1 mM TCEP Brd2BD2 (344–455), Brd3BD1 (24–146), Brd4BD1 (44–178), and Brd4BD2 (333–460), and equivalent mutant constructs were expressed with an N-terminal His6-tag in E. coli BL21 (DE3) at 18
C for 20 h using 0.2 mM IPTG. His6-tagged BDs were purified on nickel Sepharose 6 fast flow beads (GE Healthcare) by elution with increasing concentrations of imidazole. BDs were then additionally purified by sizeexclusion chromatography using a Superdex-75 column. All ¨ kta chromatography purification steps were performed using A FPLC purification systems (GE Healthcare) or glass econocolumns (Bio-Rad) at room temperature. The final purified proteins were stored in the corresponding storage buffer (see Table 1), at 80
C. 3.2

Protein Labeling

1. Dilute proteins in NHS labeling buffer to prepare 100 μL of protein to be labeled at a concentration of 10 μM (see Note 1). 2. Suspend 10 μg of RED-NHS 2nd Generation fluorophore in 25 μL DMSO and vortex to ensure proper dye suspension. The final concentration will be 600 μM.

MST and TRIC to Study PROTACs in Drug Discovery

121

3. Mix 7 μL of the RED-NHS 2nd Generation fluorophore suspension with 7 μL NHS labeling buffer to prepare 14 μL of a 300 μM dye solution in 50% DMSO. 4. Initiate the protein labeling reaction by adding 10 μL of the 300 μM dye solution to 90 μL of the 10 μM protein solution to yield 100 μL of dye–protein solution with a threefold excess of dye. Ensure the dye–protein solution is well mixed by carefully pipetting up and down several times, followed by a 5-s spin in a tabletop microfuge. 5. Incubate the labeling reaction for 30 min in the dark. 6. While the labeling reaction incubates, equilibrate the B-Column from the Monolith Protein Labeling Kit RED-NHS 2nd Generation (Amine Reactive) with the assay buffer. Remove the top cap from the B-Column and pour off the storage solution, and then remove the bottom cap. Save both caps and set aside, and then place the B-Column in a 15-mL centrifuge tube using an adapter. Fill the column with assay buffer, and allow buffer to enter the packed resin bed completely by gravity flow. Discard the flow through. Repeat filling the column with assay buffer and discarding the flow through until the column is equilibrated with 9 mL of assay buffer (see Note 2). 7. After the labeling reaction incubation is complete, transfer 100 μL of dye–protein solution to the equilibrated B-Column. Avoid contacting the inner walls of the column and load the sample directly onto the center of the resin bed. Let the sample enter the resin bed completely. 8. Add 550 μL of assay buffer to the B-column, and allow buffer to enter the resin bed completely, discarding the flow through (see Note 3). 9. Elute the labeled protein into a fresh microcentrifuge tube placed under the column by adding 450 μL of assay buffer onto the column. 10. Determine the protein concentration and degree of labeling (DOL) to verify the success of the labeling reaction. Blank your nanodrop with 2 μL assay buffer, and then record absorbance values at 280 nm and 650 nm with 2 μL labeled protein (or appropriate volume for your spectrophotometer). Use Eq. (1) to calculate the protein concentration and Eq. (2) to calculate the DOL (see Table 2). c prot ¼ DOL ¼

A prot A  A max CF ¼ 280 εprot d εprot d

c Dye A =ε A max εProt ¼ max max ¼ c Prot AProt =εProt ðA 280  A max CF Þεmax

ð1Þ ð2Þ

122

Tanja Bartoschik et al.

Table 2 Protein concentration and DOL Protein

A280

A650

Concentration

DOL

VCB

0.06

0.38

1.3 μM

1.45

BD2

Brd2

0.04

0.27

1.8 μM

0.8

Brd3BD1

0.08

0.54

2.1 μM

1.3

BD1

0.08

0.33

2.5 μM

0.7

BD2

0.06

0.14

3.9 μM

0.19

Brd4 Brd4

  εmax M1 cm1 ¼ 195 000 CF ¼ 0:04 3.3 Binary Interaction Analysis

1. Prepare 1 mM solution of MZ1 in DMSO, by adding 1 μL of MZ1 stock solution to 50 μL DMSO.

3.3.1 Protein-PROTAC Affinity Measurements with MST

2. Prepare 125 μL of MZ1 at 40 μM in assay buffer with 4% DMSO by diluting 5 μL of 1 mM MZ1 solution into 120 μL assay buffer (see Note 4). 3. Prepare five copies of a serial twofold dilution of MZ1 in a 384-well plate, each serial dilution containing 16 MZ1 concentrations. Add 10 μL assay buffer containing 4% DMSO to each well of rows B-P in columns 1–5, and then add 20 μL of MZ1 at 40 μM to row A in columns 1–5. For all columns, transfer 10 μL from well A to B and mix by pipetting up and down three times. Proceed to transfer 10 μL from well B to C, from C to D, and so on until reaching row P. After the final transfer to well P, remove 10 μL from each well in row P to finish the serial dilution with each well containing 10 μL of sample (see Note 4b). 4. Dilute the labeled proteins to 10 nM in assay buffer ensuring to make at least 200 μL (see Note 5). 5. Mix the labeled proteins with the MZ1 serial dilution. Add 10 μL of VCB to each well in column 1, using a fresh tip for each well and pipetting up and down three times to ensure good sample mixing. Repeat with the remaining proteins for columns 2–5. 6. Centrifuge the 384-well plate at room temperature for 1 min to remove any air bubbles (see Note 6). 7. Load samples from one column at a time by dipping premium coated capillaries into the microwells to aspirate the sample (see Note 7). Place the capillary with the highest concentration of MZ1 from row A on position 1 of the capillary tray and the

MST and TRIC to Study PROTACs in Drug Discovery

123

capillaries with decreasing concentrations of MZ1 from rows B to P in positions 2 to 16 of the capillary tray. Alternatively, the 384-well plate can be loaded directly into the Dianthus NT.23 instrument to conduct higher throughput measurements (see Note 8). 8. Load the sample tray in the Monolith NT.115 Pico instrument. 9. Execute a binding affinity measurement in the MO. Control Software. Set the temperature control to 25
C, the MST power to medium, and the excitation power to achieve a fluorescence of approximately 8000 counts for all capillaries in the capillary scan. Complete the fields to record the information about the target protein, ligand MZ1, buffer, and capillary type used in the assay. 3.3.2 Data Analysis

1. The MO.Control Software automatically performs a number of checks as the MST measurement is executed and data recorded. These checks include variations in fluorescence intensity, adsorption to surfaces, sample aggregation, and photobleaching rate changes. If the data pass all of these quality checks, the MO.Control Software performs an initial analysis on the data to determine the Kd of the interaction. The interaction of MZ1 with all five proteins passes all the quality checks. However, the interaction of MZ1 with Brd3BD1 and Brd4BD1 results in a biphasic dose–response curve along with a ligand-dependent fluorescence change that deserves deeper analysis (see Note 9). 2. To perform a more thorough analysis of the data and compare the interaction of MZ1 with the five proteins, analyze the acquired data in the MO.Affinity Analysis Software. Begin by creating a new MST Analysis and adding the data for the interaction of MZ1 with VCB, Brd2BD2, and Brd4BD2. Any replicates can be combined by dragging the data to the appropriate Merge Set (see Note 10). On the dose response fit tab, switch the analysis set to expert mode and change the MST evaluation strategy to manual mode and set the on-time to 1.5 s. Fit the data with a 1:1 Kd model and obtain dose– response curves that can be normalized to ΔFNorm on the Compare Results tab (see Fig. 1a). 3. To perform a deeper analysis of the biphasic dose–response seen in the MST Analysis of MZ1 interacting with Brd3BD1 and Brd4BD1, create a new Initial Fluorescence Analysis and add the data. On the dose response fit tab, set the Fluorescence Evaluation Strategy to Manual: Initial Fluorescence and compare the ligand-dependent fluorescence change at the beginning of the measurement to different time points by manually adjusting the analysis window. As the ligand-dependent fluorescence change shifts over the course of the measurement, this

124

Tanja Bartoschik et al.

Fig. 1 Binary complex formation. (a) Binary interaction of VCB (black), Brd4BD2 (dark gray), and Brd2BD2 (light gray) against PROTAC. Measurement was evaluated after 1.5 s MST on-time at medium MST power. Kd values and standard deviations of triplicate measurements are shown in the figures. (b) Binary interaction of Brd3BD1 (gray) and Brd4BD1 (black) against PROTAC. Here, the initial fluorescence was evaluated, after confirmation of binding specific ligand-induced fluorescence change using the SD-test (see sect. 3.3.3) (see Note 11). Kd values and standard deviations (n ¼ 3) are illustrated in the graphs

is most likely the cause of the biphasic dose–response observed in the MST analysis of the interaction of MZ1 with Brd3BD1 and Brd4BD1. The interaction of MZ1 with Brd3BD1 and Brd4BD1 thus needs to be analyzed based on the liganddependent initial fluorescence change (see Fig. 1b) if the fluorescence change is shown to be caused by a specific interaction between MZ1 with Brd3BD1 and Brd4BD1 through a SD-test (see Note 11). 3.3.3 SD-Test

1. To verify that the initial fluorescence change observed with the interaction of MZ1 with Brd3BD1 and Brd4BD1 is caused by an interaction and not nonspecific effects, transfer the remaining 10 μL of rows A–C and N–P (the highest three concentrations of MZ1 and lowest three concentrations of MZ1, respectively) of the Brd3BD1 and Brd4BD1 samples to nonbinding microcentrifuge tubes and centrifuge at 15,000  g for 10 min at 4
C. 2. Transfer 7 μL from the centrifuged tubes to fresh microcentrifuge tubes taking care not to disturb or transfer any pellet or aggregated sample at the bottom of the centrifuged tubes (see Note 12). 3. Add 7 μL of SD-mix to each microcentrifuge tube and mix by pipetting up and down. 4. Incubate samples for 5 min in a 95
C hot plate to ensure the proteins are fully denatured and no specific interaction between MZ1 and the protein’s structure is possible. 5. Spin samples quickly in a tabletop microfuge to ensure the samples are at the bottom of the microcentrifuge tubes.

MST and TRIC to Study PROTACs in Drug Discovery

125

6. Load samples from the microcentrifuge tubes by dipping premium coated capillaries into the sample at the bottom of the tubes (see Note 7). Place the capillaries in the capillary tray with the MZ1 decreasing from position 1 to position 6. 7. Execute a SD-test measurement in the MO.Control Software. Set the temperature control to 25
C and the excitation power to achieve a fluorescence of approximately 8000 counts for all capillaries in the capillary scan (see Fig. 2).

Fig. 2 SD-test of the binary interactions Brd4BD1 and Brd3BD1. (a) The left graph shows the initial fluorescence counts of capillary 1–3 and 14–16 of the MZ1-Brd4BD1 titration series, while the right graph represents the same samples after addition of SD-mix. (b) The left graph shows the initial fluorescence counts of capillaries 1–3 and 14–16 of the Brd3BD1-MZ1 titration series, while the right graph represents the same samples after the SD-test was carried out

126

Tanja Bartoschik et al.

3.4 Ternary Interaction Analysis 3.4.1 Ternary Complex Affinity Measurements with MST

The described assay design is one of several possible approaches to analyze ternary complex formation with MST. This assay design is ideal when the ternary interaction of different PROTACs or target proteins with a defined ubiquitin ligase is studied (see Note 13). 1. Prepare a 50 μM solution of MZ1 in DMSO through a two-step dilution by adding 1 μL of the 51 mM stock solution into 9 μL of DMSO to make a 5.1 mM solution and then adding 1 μL of this 5.1 mM solution to 101 μL of DMSO. 2. Prepare a 4 μM solution of MZ1 in assay buffer (with 8% DMSO) by adding 4 μL of the 50 μM MZ1 solution into 46 μL of assay buffer (see Note 4b). 3. Prepare 500 μL of assay buffer, supplemented with 8% DMSO, by adding 40 μL of DMSO to 460 μL of assay buffer. 4. Create a 16-point serial twofold dilution of MZ1 in a 384-well plate (see Notes 4 and 14). Transfer 50 μL of MZ1 at 4 μM to well A1, and then add 25 μL of the assay buffer supplemented with 8% DMSO to wells B1–P1. Begin the serial dilution by transferring 25 μL of solution from well A1 to well B1 and mix by pipetting up and down three times. Proceed to transfer and mix 25 μL from well B1 to C1, from C1 to D1, and so on until reaching well P1. After mixing the contents of well P1 by pipetting, finish the serial dilution by discarding 25 μL from well P1 to finish the serial dilution with 25 μL of sample in each well of column 1. 5. Prepare the four BD samples (Brd3BD1, Brd4BD1, Brd4BD2, and Brd2BD2) in a 384-well plate to form binary complexes with the MZ1 serial dilution (see Note 15). Transfer 10 μL of 100 μM solutions of Brd3BD1, Brd4BD1, Brd4BD2, and Brd2BD2 in assay buffer to wells A2–A5, respectively. Add 5 μL assay buffer to wells B2–D5 and serially dilute the BD proteins three times. In each of columns 2–5, transfer 5 μL from row A to row B and mix by pipetting up and down three times. Repeat for row B to C and for row C to D, mixing the solution after each transfer and changing pipette tips when changing columns. Discard 5 μL of solution from wells D2– D5, leaving 5 μL of BD sample in wells A2–D5. Finally, prepare 70 μL of each BD protein at 8 μM and transfer 5 μL of the Brd3BD1, Brd4BD1, Brd4BD2, and Brd2BD2 solutions to the remaining wells of columns 2–5, respectively. 6. Combine the MZ1 serial dilution with the BD proteins to form binary complexes of the MZ1 with the four BD proteins, separately. Transfer 5 μL of the MZ1 solution in tube 1 to each of wells A2–A5 and mix by pipetting up and down three times, taking care to avoid air bubbles (see Note 6). Use a fresh pipette tip for each well. Proceed to transfer 5 μL of the MZ1

MST and TRIC to Study PROTACs in Drug Discovery

127

serial dilution to wells B2–B5 from tube 2 to C2–C5 from tube 3, continuing down the plate to P2–P5 from tube 16. 7. Incubate the interaction for 10 min at room temperature. Seal the plate to avoid any evaporation. 8. While the plate incubates, prepare 800 μL of a 10 nM solution of the RED-NHS labeled VCB prepared in section 3.2 using assay buffer as the diluent. 9. After the 10 min incubation, add 10 μL of the 10 nM RED-NHS labeled VCB solution to each well containing the MZ1-BD mixture (wells A2–P5) and mix by pipetting up and down three times. 10. Centrifuge the plate for 1 min at room temperature to remove any air bubbles. 11. Load samples from one column at a time, starting with column 2, by dipping one premium coated glass capillary per well to aspirate the sample (see Note 7). For each BD, place the capillary with the highest concentration of MZ1 from row A on position 1 of the capillary tray and the capillaries with decreasing concentrations of MZ1 from rows B to P in positions 2 to 16 of the capillary tray. Alternatively, the 384-well plate can be loaded directly into the Dianthus NT.23 instrument to conduct higher throughput measurements (see Note 8). 12. Load the sample tray in the Monolith NT.115Pico instrument. 13. Execute a binding affinity measurement in the MO.Control Software. Set the temperature control to 25
C, the MST power to medium, and the excitation power to 5% (see Note 16). Complete the fields to record the information about the target protein, ligand MZ1-BD complex, buffer, and capillary type used in the assay. 3.4.2 Data Analysis

1. The MO.Control Software automatically checks for variations in fluorescence intensity, sample adsorption to capillary walls, sample aggregation, and photobleaching rate changes. If the data pass all of these quality checks, the MO.Control Software determines the Kd of the interaction. All four ternary complex formation measurements pass the quality checks performed by the MO.Control Software (see Note 17). 2. Open the acquired data in the MO.Affinity Analysis Software for data processing and analysis. Begin by creating a new MST Analysis and adding the data for the interaction of VCB with the Brd3BD1-MZ1 complex. Any replicates can be combined by dragging the data to the appropriate Merge Set (see Note 10). On the Dose Response Fit tab, switch the analysis set to expert mode and change the MST evaluation strategy to manual mode

128

Tanja Bartoschik et al.

Fig. 3 Determination of binding affinity of ternary interactions of VCB with MZ1-BD complexes. Measurements with Brd3BD1 (a) and Brd4BD1 (c) are evaluated after 1.5 s MST on-time. Brd4BD2 (b) and Brd2BD2 (d) are evaluated after 20 s MST on-time. Kd values are shown in each figure. Error bars indicate the standard deviation of three independent measurements (n ¼ 3)

and set the on-time to 1.5 s (see Note 18). Fit the data with a 1:1 Kd model and obtain a dose–response curve that can be normalized to FNorm on the Compare Results tab (see Fig. 3). Repeat with the data for the interaction of VCB with the Brd4BD1-MZ1, Brd4BD2-MZ1, and Brd2BD2-MZ1 complexes, setting the on-time to 1.5 s for the Brd4BD1-MZ1 complex and 20 s for the Brd4BD2-MZ1 and Brd2BD2-MZ1 complexes. 3.4.3 Cooperativity Assessment of Ternary Complex

1. Calculate the cooperativity (α) of the ternary complex formation, an important metric related to the successful ubiquitinylation of the target protein by the ubiquitin ligase [33]. Divide the Kd of the binary interaction between MZ1 and VCB determined in section 3.3 by the Kd of the ternary interaction between MZ1 bound to the BD proteins and VCB determined in section 3.4 (see Table 3). Higher cooperativity values indicate an attractive interaction between the ubiquitin ligase and target protein catalyzed by the PROTAC MZ1.

MST and TRIC to Study PROTACs in Drug Discovery

129

Table 3 Cooperativity values BD

Brd4BD2

Brd2BD2

Brd3BD1

Brd4BD1

α

473

54.6

42.4

22.7

Table 3 lists the cooperativity values α for each BD. The values are calculated by dividing the Kd of the binary interaction between MZ1 and VCB by the Kds for the ternary interaction between MZ1 bound to a BD and VCB.

4

Notes 1. For proteins with a concentration of less than 100 μM or proteins stored in Tris buffers, it is best to perform a buffer exchange into the NHS labeling buffer using desalting A-Column in the Monolith Protein Labeling Kit RED-NHS 2nd Generation as dilution into the NHS labeling buffer will leave too much of the original buffer components that can have a negative effect on the protein labeling efficiency. 2. If the equilibration of the B-column is finished before the 30 min labeling incubation is completed, place the caps you set aside back on the column to prevent the resin bed from drying out. 3. Use care when adding the 550 μL of assay buffer to the B-column to avoid disturbing the resin bed and disturbing the separation of the labeled protein from the unreacted dye. 4. (a) Serial dilution of the ligand should begin at a concentration approximately 50-fold the estimated Kd of the protein–ligand interaction to ensure saturation of the protein target is achieved and the full dose–response curve is captured. (b) The final DMSO concentration should always be kept as low as possible (best below 5%) to ensure that the protein is not affected in its activity. In general, measurements at very high additive/solvent concentrations (>2% DMSO, >5% glycerol, >100 mM sucrose) are possible without limitations when applied at constant concentrations throughout the dilution series. 5. For proper dissociation constant determination, the target protein concentration must be below the Kd (e.g., for an interaction with a Kd of 100 nM, the concentration of the target protein should be no greater than 100 nM). 6. Failure to remove air bubbles and aggregated sample from solution can affect the quality of the results. Both can be

130

Tanja Bartoschik et al.

removed by centrifugation of the samples at 15,000  g for 10 min at 4
C. 7. Aspiration of samples into capillaries is best performed horizontally to eliminate the force of gravity from counteracting the capillary force. One of the easier approaches is to hold the 384-well plate vertically or place the plate in the NanoTemper plate holder and then dip the capillaries into the plate column with the samples to be measured. 8. Dianthus instruments leverage the Temperature-Related Intensity Change (TRIC) component of the MST signal to enable measurements in industry standard microwell plates. TRIC is based on the generation of a rapid and highly precise temperature change in a sample well by infrared (IR) laser light. Changes in sample fluorescence upon activation of the IR laser are monitored to characterize interactions and derive affinity constants. 9. Any measurement that results in a biphasic dose–response curve from an MST analysis and also has a ligand-dependent change in initial fluorescence needs further analysis as the biphasic dose–response from the MST analysis can be an artifact from a time or temperature-related shift to the initial fluorescence change. Best practice is to perform an initial fluorescence analysis in MO.Affinity Analysis Software with the Fluorescence Evaluation Strategy set to Manual: Initial Fluorescence and to compare the ligand-dependent fluorescence change at various points in the measurement by manually adjusting the analysis window. If a time-dependent shift is observed for the initial fluorescence change, this is most likely the cause of the biphasic dose–response seen in the MST analysis. 10. We selected to perform our assays in triplicate. Like all experiments, repeating the assay increases the statistical significance of the results. 11. The SD-test is not suitable if the target is a fluorescent fusion protein like GFP or YFP. These fluorescent proteins will be denatured as well, and no fluorescence will be left for analysis. If potassium salts are used in the assay buffer, SDS should be avoided due to precipitation of the salt. Instead, a final concentration of 4 M urea should be used to denature the proteins to execute the SD-test. For samples containing RED-tris-NTA labeled protein, please perform an ECP-Test instead of an SD-test [35]. 12. It is essential to ensure that none of the pellet after centrifuging is transferred to the new tubes. If the pellet is disturbed, centrifuge again for at least 10 min at 15,000  g.

MST and TRIC to Study PROTACs in Drug Discovery

131

13. While the described assay design is ideal to study the ternary interaction of different PROTACs and target proteins with a defined ubiquitin ligase while in addition eliminating the hook effect (the hook effect describes how high concentrations of a bifunctional molecule can prevent ternary complex formation), alternative assay designs are possible that offer different advantages. By fluorescently labeling the target protein instead of the ubiquitin ligase, the assay can be easily modified to study the ternary interaction of different PROTACs and ubiquitin ligases with a single target protein while still avoiding the hook effect. Pushing the assay design further, unpublished results have shown assays utilizing excess PROTAC preincubated with the fluorescently labeled protein to study ternary interactions are able to recapitulate the cooperativity ranking as the assay design described in this chapter despite the hook effect. When designing an assay using an excess of PROTAC preincubated with the fluorescently labeled protein, care must be taken to strike a balance between having enough but not too much PROTAC in solution. While sufficient PROTAC is necessary to ensure saturation of the fluorescently labeled protein to enable ternary complex formation, a too large excess of PROTAC will lead to the hook effect, precluding the target protein and ubiquitin ligase finding the same PROTAC to form a ternary complex. 14. The use of low-binding plastics (multiwell plates, microcentrifuge tubes, and pipette tips) is recommended to prevent loss of sample through sticking to plastics. Losses of small amounts of sample that go unnoticed when working at higher concentrations are more pronounced and can have a larger effect when working at the low concentrations common in MST measurements. 15. It is important that the concentration of BD protein is high enough to ensure 95% of the MZ1 in solution is bound to the BD. Any MZ1 not bound to BD will interact with VCB without any cooperativity effects. Having two different ligands in solution (MZ1 in complex with BD and MZ1 alone) can mask the effects of cooperativity on the ternary complex formation. Given the Kds of the binary interactions of MZ1 with the BD proteins, the concentration of BDs is set so 95% of the MZ1 is bound by a BD at all concentrations of the MZ1 serial dilution. 16. In the MO.Control Software, the LED power can be adjusted automatically by selecting the “auto” function on the plan page of the software. 17. Of all the quality checks performed by the MO.Control Software, the initial fluorescence (the fluorescence of the samples

132

Tanja Bartoschik et al.

before the IR laser is switched on) is one of the most important. Several of the most frequently encountered complications that impact MST assays can be identified and diagnosed via an analysis of the initial fluorescence. As the concentration of the fluorescent target is constant in an MST assay, the initial fluorescence should be homogenous for all ligand concentrations. While random fluctuations in the initial fluorescence typically signify protein aggregation that can be addressed through alterations to buffer composition or pH, ligand-dependent changes require a SD-Test (for covalently fluorescently labeled targets) or ECP-Test (for his-tag fluorescently labeled targets) to distinguish between interaction-based effects and artifacts such as adsorption to labware or ligand autofluorescence. 18. It is best to study interactions using the lowest MST power and shortest analysis time that resolves a good amplitude (typically 5 or more FNorm units) and signal to noise ratio (typically 13 or more). For most interactions, this is the medium MST power. References 1. An S, Fu L (2018) EBioMedicine smallmolecule PROTACs: An emerging and promising approach for the development of targeted therapy drugs. EBioMedicine 36:553–562 2. Mishra SK et al (2018) SMMDB: a web-accessible database for small molecule modulators and their targets involved in neurological diseases. Database 2018:1–12 3. Lavanya V, Adil M, Ahmed N, Rishi AK, Jamal S (2014) Small molecule inhibitors as emerging cancer therapeutics. Integr Cancer Sci Ther 1:39–46 4. Markossian S, Ang KK, Wilson CG, Arkin MR (2018) Small-molecule screening for genetic diseases. Annu Rev Genomics Hum Genet 19:263–288 5. Sun X et al (2019) PROTACs: great opportunities for academia and industry. Signal Transduct Target Ther 4:64 6. Toure M, Crews CM (2016) Small-molecule PROTACS: new approaches to protein degradation. Angew Chem Int Ed Engl 55:1966–1973 7. Wells JA, McClendon CL (2007) Reaching for high-hanging fruit in drug discovery at protein-protein interfaces. Nature 450:1001–1009 8. Dang CV, Reddy EP, Shokat KM, Soucek L (2017) Drugging the ‘undruggable’ cancer targets. Physiol Behav 176:139–148

9. Sun X, Rao Y (2019) PROTACs as potential therapeutic agents for cancer drug resistance. Biochemistry. https://doi.org/10.1021/acs. biochem.9b00848 10. Burslem GM et al (2018) The Advantages of Targeted Protein Degradation Over Inhibition: An RTK Case Study. Cell Chem Biol 25:67–77.e3 11. Gu S, Cui D, Chen X, Xiong X, Zhao Y (2018) PROTACs: An emerging targeting technique for protein degradation in drug discovery. BioEssays 40:e1700247 12. Leestemaker Y et al (2017) Proteasome activation by small molecules. Cell Chem Biol 24:725–736.e7 13. Mullard A (2019) First targeted protein degrader hits the clinic. Nat Rev Drug Discov. https://doi.org/10.1038/d41573-01900043-6 14. Wang Y, Jiang X, Feng F, Liu W, Sun H (2020) Degradation of proteins by PROTACs and other strategies. Acta Pharm Sin B 10:207–238 15. Benchekroun M (2019) The advent of directed protein degraders in drug discovery. Futur Drug Discov 1:FDD16 16. Schiebel J et al (2015) One Question, Multiple Answers: Biochemical and biophysical screening methods retrieve deviating fragment hit lists. ChemMedChem 10:1511–1521 17. Schiebel J et al (2016) Six biophysical screening methods miss a large proportion of

MST and TRIC to Study PROTACs in Drug Discovery Crystallographically discovered fragment hits: a case study. ACS Chem Biol 11:1693–1701 18. Rainard JM, Pandarakalam GC, McElroy SP (2018) Using Microscale thermophoresis to characterize hits from high-throughput screening: a European Lead factory perspective. SLAS Discov 23:225–241 19. Jose´-Ene´riz ES et al (2017) Discovery of firstin-class reversible dual small molecule inhibitors against G9a and DNMTs in hematological malignancies. Nat Commun 8 20. Pollack SJ et al (2011) A comparative study of fragment screening methods on the p38α kinase: new methods, new insights. J Comput Aided Mol Des 25:677–687 21. Martin LJ et al (2016) Structure-based design of an in vivo active selective BRD9 inhibitor. J Med Chem 59:4462–4475 22. Linke P et al (2016) An automated Microscale thermophoresis screening approach for fragment-based Lead discovery. J Biomol Screen 21:414–421 23. Jerabek-Willemsen M et al (2014) MicroScale thermophoresis: interaction analysis and beyond. J Mol Struct 1077:101–113 24. Li W et al (2017) Benserazide, a dopadecarboxylase inhibitor, suppresses tumor growth by targeting hexokinase 2. J Exp Clin Cancer Res 36:1–12 25. Joanne L Parker, Simon Newstead. Molecular basis of nitrate uptake by the plant nitrate transporter NRT1.1. Nature 507, 68–72 (2014) 26. Raj I et al (2017) Structural basis of egg coatsperm recognition at fertilization. Cell 169:1315–1326.e17 27. Gadd MS et al (2017) Structural basis of PROTAC cooperative recognition for selective protein degradation accession codes atomic coordinates and structure factors for hsBrd4

133

BD2-MZ1-hsVHL-hsEloC-hsEloB have been deposited in the protein data Bank (PDB) under accession number. Nat Chem Biol 13:514–521 28. Bondeson DP et al (2019) Lessons in PROTAC design from selective degradation with a promiscuous warhead. Cell Chem Biol 25:78–87 29. Zengerle M, Chan KH, Ciulli A (2015) Selective small molecule induced degradation of the BET Bromodomain protein BRD4. ACS Chem Biol 10:1770–1777 30. Testa A et al (2018) 3-Fluoro-4-hydroxyprolines: synthesis, conformational analysis, and Stereoselective recognition by the VHL E3 ubiquitin ligase for targeted protein degradation. J Am Chem Soc 140:9299–9313 31. Han X et al (2019) Discovery of highly potent and efficient PROTAC degraders of androgen receptor (AR) by employing weak binding affinity VHL E3 ligase ligands. J Med Chem 62:11218–11231 32. Farnaby W et al (2019) BAF complex vulnerabilities in cancer demonstrated via structurebased PROTAC design. Nat Chem Biol 15:672–680 33. Roy MJ et al (2019) SPR-measured dissociation kinetics of PROTAC ternary complexes influence target degradation rate. ACS Chem Biol 14:361–368 34. Beveridge, R. et al. (2019) Native mass spectrometry can effectively predict PROTAC efficacy. bioRxiv 851980. https://doi.org/10. 1101/851980 35. Bartoschik T et al (2018) Near-native, site-specific and purification-free protein labeling for quantitative protein interaction analysis by MicroScale thermophoresis. Sci Rep 8:4977

Chapter 7 An In Vitro Pull-down Assay of the E3 Ligase:PROTAC: Substrate Ternary Complex to Identify Effective PROTACs Daniel P. Bondeson, Blake E. Smith, and Alexandru D. Buhimschi Abstract Assessing the specificity of PROTACs and confirming their proposed mechanism of action are critical for a robust targeted protein degradation program. Owing to their novel mechanism, new assays are needed to meet these goals. We and others have shown that a common explanation of PROTAC efficacy is the ability of the PROTAC to form a ternary complex between the E3 ubiquitin ligase and the target protein. In this chapter, we provide a simple in vitro method to quickly and inexpensively assess this property of PROTAC molecules. We provide detailed instructions for the purification of the specific E3 ubiquitin ligase VHL and then a generic protocol which can be adapted to any E3 ligase and substrate protein combination. This accessible method to study the ternary complex can strengthen any PROTAC-focused medicinal chemistry effort. Key words PROTAC , VHL, E3 ubiquitin ligase, Ternary complex, Degrader, Targeted protein degradation

1

Introduction Recently, targeted protein degradation has emerged as an exciting opportunity for drug discovery. Whereas traditional occupancybased inhibitors are limited to enzymes with readily tractable active sites, targeted protein degradation has potential to drug the ‘undruggable’ proteome. While this field has existed since 2001, landmark publications in 2015 and onward have demonstrated the broad applicability of small molecules which hijack the endogenous ubiquitin–proteasome system to induce the specific degradation of proteins [1–5]. These small molecules, called proteolysis-targeting chimeras (PROTACs), are heterobifunctional compounds with distinct binding moieties for an E3 ubiquitin ligase and the substrate to be degraded, connected by a chemical linker. A second class of small molecules used in targeted protein degradation is ‘molecular glues,’ such as the immunomodulatory imide drugs (IMiDs), which remodel the surface of the E3 ubiquitin ligase cereblon

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_7, © Springer Science+Business Media, LLC, part of Springer Nature 2021

135

136

Daniel P. Bondeson et al.

(CRBN) to recruit and cause the degradation of numerous and otherwise unamenable ‘neo-substrates’ [6–8]. Interestingly, molecular glues may be a more general phenomenon, not limited to the IMiD-reprogramming of CRBN [9–11]. From these descriptions, it is clear that traditional assays used in medicinal chemistry programs to identify and improve enzymatic inhibitors will not directly translate to targeted protein degradation. For example, if an enzymatic inhibitor were being developed, one would seek to develop high-throughput binding and enzymatic assays to confirm that the inhibitor was indeed binding and inhibiting the catalytic activity of the enzyme [12, 13]. With PROTACs, however, the binding and enzymatic requirements are less straightforward [14]. Arising from this complexity, the main screening method used to identify degraders is in cellula degradation via immunoblotting or higher-throughput methods [15, 16]. While in cellula protein degradation or loss of cancer cell viability are the typical goals of PROTAC molecules, these assays do not provide the mechanistic insights required for a robust targeted protein degradation program. For example, if there are multiple mechanisms by which a protein is being degraded in cells, understanding the structure–activity relationship of a given compound will be challenging [17, 18]. A recent study demonstrated that small structural changes to a PROTAC abrogated the desired chimeric activity and converted the molecule into a molecular glue, recruiting a different substrate protein altogether and confounding the cellular toxicity readout [19]. Another theoretical possibility would be ‘bystander’ or ‘collateral’ degradation, in which the target protein is within a larger protein complex, and other members of the complex are erroneously ubiquitinated. Finally, appending a linker and E3 ligase ligand to the targeting warhead has resulted in compounds which both recruit an E3 ligase and also destabilize the target protein with a non-PROTAC mechanism [2]. In each of these examples, constructing a robust structure–activity relationship would be challenging if only degradation or cell viability measurements were taken. Rather than studying the degradation of a target protein, we describe here a medium-throughput assay which more directly addresses a fundamental prerequisite for degrader action. Before a target protein can be degraded or ubiquitinated, the PROTAC must induce proximity with the E3 ligase. This step—formation of a ternary complex or trimer between E3 ligase:PROTAC:substrate—has recently been highlighted as the most discriminating feature yet identified between functional and nonfunctional PROTACs [20, 21]. Small chemical modifications to PROTACs can cause drastic changes in the ternary complex binding equilibria, which then impact ubiquitination and degradation outcomes [22]. Although we expect that there will be examples of high-affinity ternary complexes which do not result in

Assessing the PROTAC-Induced Ternary Complex

137

degradation—for example, when there are no accessible residues for ubiquitination—we expect that most PROTAC programs can be explained by ternary complex formation. For these reasons, we propose that understanding the ternary complex will be a key component to developing a PROTAC medicinal chemistry program or to confirm that the mechanism of action of a particular compound is via a genuine PROTAC mechanism. Furthermore, although an in vitro ubiquitination assay provides alternative information (e.g., which lysine residues are accessible for ubiquitination), the protein purification requirements for assessing the PROTAC-induced ternary complex are far more accessible for standard molecular biology laboratories [23]. Many experimental approaches give information on the PROTAC-induced ternary complex with differing levels of detail, throughput, and technical caveats. At one end of the spectrum are structural approaches, such as X-ray crystallography [23, 24]. While this provides the most in-depth information on the interactions between PROTAC, substrate, and E3 ligase, it is very low throughput. Robust biophysical measurements can also be calculated from isothermal titration calorimetry or surface plasmon resonance, but these methods require specific instrumentation and suffer from low-throughput methodology. On the high-throughput end of the spectrum are plate-based assays, most commonly the beadbased AlphaLISA proximity experiment [3, 20, 25]. However, this biophysical assay can be plagued by false-negatives. For example, bead-binding tags must be fused to either terminus of both E3 ligase and substrate protein, and at times, no combination of fusion termini yields a functional assay. Furthermore, all of these experiments require expensive and specialized equipment. Here, we outline a medium-throughput assay which uses equipment found in most molecular biology laboratories. This method is based on the precipitation of a target protein in the presence of a PROTAC molecule and E3 ligase-coated sepharose beads. This assay, which we call the Trimer Pull-down Assay (TPA), can be readily established with minimal troubleshooting [20, 22]. After purification of the E3 ligase and substrate protein, the ternary complex formation of several different compounds can be evaluated in parallel in a five-hour assay (see Fig. 1). This protocol involves coating sepharose beads with the E3 ligase, incubating the beads with the substrate protein and PROTAC, washing away unbound protein, and finally assessing the amount of ternary complex formed via western blot.

138

Daniel P. Bondeson et al.

Fig. 1 Schematic representation of the Trimer Pull-down Assay. In this four-step assay, (1) GST-VBC is immobilized on glutathione sepharose 4B beads, (2) incubated in the presence of substrate protein and test compounds (cmpd 1 or cmpd 2) or vehicle (
), (3) washed extensively to remove unbound substrate protein, eluted, and (4) run on western blot to analyze ternary complex formation. In this simulated western blot, we demonstrate a situation where two compounds are able to form the ternary complex, but compound 1 more efficiently forms the complex (e.g., the complex has a higher affinity and/or is more stable)

2

Materials

2.1 GST-VBC Protein Expression

1. Plasmids for bacterial expression of GST-VHL, Elongin B, and Elongin C. We recommend purchasing the following plasmids from Addgene: pACYC-Duet1-Elongin B/C (Addgene Catalog #110274) and VHL-pGex2TK (Addgene Catalog #20790). 2. Codon-optimized E. coli, such as BL21(DE3)-RIPL. 3. Sterile Luria-Bertani (LB) liquid culture medium: 1% NaCl, 1% Tryptone, 0.5% Yeast Extract, pH 7.0.

Assessing the PROTAC-Induced Ternary Complex

139

4. Kanamycin, Carbenicillin, and Chloramphenicol antibiotics. 5. Isopropyl β-D-1-thiogalactopyranoside (IPTG). 6. Lysis Buffer: 30 mM Tris (pH 8.0), 200 mM NaCl, 5 mM DTT, 5% glycerol, 1 EDTA-free Protease Inhibitor Cocktail. 7. Elution Buffer: 50 mM Tris (pH 8.0), 200 mM NaCl, 1 mM DTT, 10 mM glutathione. 2.2 GST-VBC Protein Purification

1. Glutathione sepharose 4B beads. 2. Ion Exchange Buffer A: 30 mM Tris (pH 8.0), 5% glycerol, 1 mM DTT. 3. Ion Exchange Buffer B: 30 mM Tris (pH 8.0), 5% glycerol, 1 mM DTT, 1 M NaCl. 4. Fast protein liquid chromatography (FPLC) system. 5. Anion exchange column (e.g., Mono Q 5/50 GL). 6. Protein concentrators with a 50 kDa molecular weight cutoff (MWCO). 7. Size exclusion chromatography column (e.g., HiLoad Superdex 200). 8. Size Exclusion Buffer: 30 mM Tris (pH 8.0), 10% glycerol, 1 mM DTT, 100 mM NaCl. 9. Poly-Prep gravity flow column.

2.3 In Vitro Trimer Pull-down Assay (TPA)

1. Substrate protein (see Subheading 3.3). 2. TPA Blocking Buffer: 30 mM Tris (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 2 mM DTT, 100 mg/mL Bovine Serum Albumin, 10% glycerol, 0.01% NP-40. 3. TPA Wash Buffer: 30 mM Tris (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 2 mM DTT, 0.1 mg/mL Bovine Serum Albumin, 10% glycerol, 0.01% NP-40. 4. PROTAC, control compound(s), Dimethylsulfoxide (DMSO), and (3-((3-cholamidopropyl) dimethylammonio)-1-propanesulfonate) CHAPS detergent. 5. Antibodies for VHL (Cell Signaling Technology Catalog #68547) and substrate protein(s). 6. 4 SDS sample buffer containing 5% β-mercaptoethanol.

2.4 Common Laboratory Equipment and Supplies

1. 37  C incubator with shaker for bacterial cultures. 2. Ultracentrifuge for harvesting bacteria. 3. Autoclave. 4. Cold room. 5. Liquid nitrogen. 6. 12% SDS-PAGE gels.

140

Daniel P. Bondeson et al.

7. Standard western blotting supplies. 8. Microfluidizer and/or Branson digital sonifier. 9. 200 μL, 1.5 mL, 15 mL, 50 mL, 250 mL tubes. 10. Microcentrifuge. 11. Erlenmeyer flasks suitable for protein expression. 12. Ponceau S stain. 13. TBS-Tween 20 (TBS-T) solution. 14. Coomassie stain.

3 3.1

Methods Timing

3.2 GST-VBC Purification

3.2.1 GST-VBC Protein Expression (4-Day Procedure)

l

The VHL complex purification protocol takes several days, with many potential stopping points.

l

4 days with minimal effort to generate pellets of bacteria expressing the VHL complex.

l

2–3 days of focused effort to purify the protein from bacterial lysate, including GST pull-down, anion exchange chromatography, and size exclusion chromatography.

l

The TPA takes about 1 day to perform and a second day to analyze the results via western blot.

Below, we describe the expression and purification of the von Hippel–Lindau (VHL) substrate-binding domain (along with cofactors Elongin B and C) fused to a glutathione-s-transferase (GST) tag which can be directly used in the TPA (Subheading 3.3) [20, 22]. We refer to this complex herein as ‘GST-VBC.’ VHL has been extensively utilized by PROTACs, but other E3 ligases can also be purified using a similar protocol. The protein expression (Subheading 3.2.1) likely needs to be optimized for each protein (e.g., CRBN protein typically requires expression in insect cells), but the purification (Subheadings 3.2.2–3.2.4) is fairly representative of most protocols. Perform all steps of the purification at 4  C unless indicated otherwise. Be sure to chill all buffers to 4  C, use properly sterilized glassware, and ensure proper sterile technique for all bacterial culture preparation. 1. Day 1: Co-transform suitable BL21 E. coli (e.g., codonoptimized (DE3)-RIPL cells) with plasmids (see Subheading 2.1.1) and plate on selection LB agar plates containing appropriate antibiotics. Grow in 37  C incubator overnight. Expect slower growth of bacteria due to multiple selection markers,

Assessing the PROTAC-Induced Ternary Complex

141

and remember that codon-optimized E. coli often contain selection markers themselves. 2. Day 2: Pick one colony and inoculate a 50 mL ‘starter culture’ containing appropriate antibiotics. Grow in 37  C shaking incubator overnight, again recognizing the slower growth of multiply-selected bacteria (see Note 1). 3. Day 3: Dilute the Day 2 starter culture 200-fold into the final expression volume of LB with appropriate antibiotics (see Note 2). 4. Grow the bacteria at 37  C with shaking, monitoring the OD600 until the culture reaches OD600 ¼ 0.8–1.0, at which point chill the culture to 16  C. 5. Induce expression with IPTG at a final concentration of 0.4 mM, and return the culture to the shaking incubator. Incubate at 16  C overnight (14–16 h). 6. Day 4: Spin bacterial culture in clean centrifuge tubes at 3000 rcf for 10 min at 4  C to collect cell pellets. Dispose of the LB medium by decanting or aspiration. Continue to protein purification (step 2 below) or flash-freeze the cell pellet in liquid nitrogen and store at
80  C for future use. 3.2.2 Cell Lysis and GST Pull-down (1-Day Procedure)

1. Day 1: Resuspend the pellet in chilled Lysis Buffer until homogenous. Ensure that protease inhibitors are present at an appropriate concentration. 2. For complete bacterial cell lysis, we recommend one of the following strategies: (a) For microfluidization: Pass the homogenized cells through a microfluidizer three times at 15,000 PSI. (b) For sonication: Sonicate the homogenized cells four times on ice using a Branson digital sonifier (settings: 50% amplitude, 0.5 s on/0.5 s off, total sonication time of 1.5 min each round, with 5 min rest between each sonication cycle to prevent overheating). 3. After lysis, clarify the cell lysate by centrifugation at 20,000 rcf for 1 h at 4  C. 4. During centrifugation, prepare glutathione sepharose 4B beads by washing twice in water, followed by two washes in Lysis Buffer (see Note 3). To wash, spin the beads at 150 rcf, decant the supernatant, and resuspend in five column volumes of Lysis Buffer. 5. After centrifugation, decant the lysate into a new clean vessel and add the prewashed glutathione sepharose 4B beads to the lysate. Incubate with gentle rotation at 4  C for 2–3 h.

142

Daniel P. Bondeson et al.

6. During the last hour of this incubation, prepare the Poly-Prep gravity flow column by washing with a small volume of Lysis Buffer at 4  C. Ensure that the column always remains in buffer and does not run dry. 7. After 2–3 h of incubation, transfer the lysate/bead mixture to the equilibrated Poly-Prep column and collect the flowthrough. 8. Wash the beads by adding five column volumes of Lysis Buffer, sealing the Poly-Prep column, and incubating with gentle rotation for 10 min. Allow the column to drain into a new tube, and repeat the wash step until there is no protein eluting off the column (see Note 4). 9. Elute GST-VBC from the beads by adding one column volume of Elution Buffer, sealing the Poly-Prep column, and incubating with gentle rotation for 10 min. Allow the column to drain into a new tube, and repeat this step until the elution fractions are free of protein. 10. Run a 12% SDS-PAGE gel, and assess protein identity and purity via coomassie staining (see Note 5). Pool all elution fractions containing GST-VBC. 3.2.3 Anion Exchange Purification (Half-Day Procedure)

1. Dilute the sample to reduce the salt concentration to 98% purity) should be used to prepare PBS-T with 5% BSA to reduce background signal. 6. Pipette onto the side of the wells to minimize well to well variability.

196

Karteek Kadimisetty et al.

5 Use of TUBEs as Capture Reagents for Detection of Protein Polyubiquitination from Biological Matrices Detecting quantitative changes in the ubiquitination levels of a specific substrate protein in response to external stimuli, e.g., stress, cytokine exposure, and drug candidate treatment, is challenging and requires a less complex yet sensitive method. Current methods of immunoprecipitation followed by gel electrophoresis and Western blot analysis are long, labor-intensive, low throughput, and only semiquantitative at best. LifeSensors’ UbiQuant™ S is a facile, robust, and quantitative alternative to IP/WB analysis intended for the relative determination of the concentration of a specific ubiquitinated, target protein in cellular lysates. The UbiQuant™ S is a sandwich ELISA assay in which total ubiquitinated proteins in a cell lysate are captured in the wells of a precoated microtiter plate using TUBEs. Unbound protein is removed by washing and then the amount of bound target protein is determined using an antibody specific to either the target protein or an epitope tag incorporated into the target protein. After removing the unbound antibody fraction, the bound antibody fraction is measured using an enzyme-linked anti-antibody and detected by chemiluminescence [13]. The kit from LifeSensors provides a choice of antibodies to well-characterized epitope tags, e.g., myc, HA, V5, GST, Flag®. The assay can be used for measuring changes in substrate polyubiquitination resulting from up- or down-regulation of the activity of the E1 activating enzyme, E2 conjugating enzymes, E3 ligases, or DUBs [14]. More important, the UbiQuant S platform is able to relatively measure the induced polyubiquitination and subsequent degradation of target proteins by a PROTAC® molecule (Fig. 4). To demonstrate the utility of this assay, a protocol using tagged, transfected substrate proteins and their cognate E3s is described. 5.1

Materials

Prepare all solutions using ultrapure water (prepared by purifying deionized water, to attain a sensitivity of 18 MΩ cm at 25
C) and analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise). Diligently follow all waste disposal regulations when disposing waste materials. 1. 1 PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2PO4, 1.8 mM KH2PO4 pH 7.4. Filter using a 0.22 μm membrane. 2. PBS-T: 0.1% Tween-20 in 1 PBS. 3. Cell Lysis Buffer: 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 10% glycerol. Store at 4
C. Prior to use, add the following inhibitors to attain indicated final concentration: 50 μM PR-619, 5 mM 1,10-phenanthroline (o-PA), 2 mM PMSF, 25 μM MG132, Protease Inhibitor Cocktail.

Tandem Ubiquitin Binding Entities (TUBEs)

197

Fig. 4 Validation of the UbiQuant S assay for PROTAC® molecules screening. (a) Jurkat cells were treated with dBET1 PROTAC® (1 μM) at different time points. Cell lysates (15 μg/well) were tested as described in the protocol. Results are expressed as fold change of average chemiluminescent intensity  SD (n ¼ 3). (b) Same cell lysates (25 μg) as in (a) were immunoblotted for the detection of BRD3 and β-actin. Treatment of cells with the proteasome inhibitor MG132 was used a control 5.2

Methods

1. For cells grown at 80% confluence in 10 cm2 dishes (select treated samples and control samples appropriately), add 500 μL of Cell Lysis Buffer. 2. Collect cells by scraping and transfer the lysate to 1.5 mL centrifugation tube. 3. Clarify lysate by high speed centrifugation at 14,000  g for 10 min at 4
C. Determine the protein concentration of each sample. Proceed with the dilution of the lysate for the assay. 4. Dilute test samples to a concentration between 25 and 400 μg/ mL using PBS. The final volume will depend on the number of replicates for each sample. Alternatively, perform a serial twofold dilution of each test sample. 5. Pipette 100 μL each of: blank, controls, and test samples into appropriate wells and cover with plastic plate sealer. Incubate for 1 h at room temperature. 6. Wash plate four times with ~250 μL/well of PBS-T using a multichannel pipette or an automatic plate washer. After the last wash, remove the last droplets of buffer by lightly tapping the plate (upside down) on paper towels or other blotting paper. DO NOT ALLOW WELLS TO DRY COMPLETELY. 7. Add 100 μL of diluted primary antibody to each well, cover with plastic plate seal, and incubate for 1 h at room temperature. 8. Repeat the washing process. 9. Add 100 μL of diluted enzyme labeled detection antibody to each well, cover with plate seal, and incubate for 1 h at room temperature.

198

Karteek Kadimisetty et al.

10. Repeat the washing process. 11. Just before detecting, mix 150 μL of detection reagent 1 (Luminol) and 150 μL of detection reagent 2 (hydrogen peroxide) into 10 mL of water. Add 100 μL of this detection reagent to each well. Incubate for 2 min and read in a plate reader optimized for chemiluminescence. Chemiluminescence can be measured for up to 20 min after addition of the luminescence reagent. 12. Calculate the mean chemiluminescent counts for blank, controls, and unknowns. Subtract the mean counts for the blank from each sample. 13. The data can be plotted as a bar graph or as a percent of the maximum value. If performing a compound dose-response assay, the maximum value should correspond to a zero concentration dose and the data can be fit to a sigmoidal plot to determine the ED50. 5.3

Notes

1. The amount of cell lysate per well must be determined appropriately per the abundance of the target protein. 2. Appropriate treatments such as proteasome inhibitor (MG-132) and DUB inhibitor (PR-619) are recommended to monitor polyubiquitination signal. 3. For detection of polyubiquitination of low-level proteins, LifeSensors has developed the UbiQuant Ultra kit.

6

Imaging of Polyubiquitinated Proteins Using Fluorescent-Tagged TUBEs The imaging protocol for polyubiquitinated proteins follows a general methodology for immunostaining [15]. TUBEs that are labeled with a fluorescent tag such as TAMRA (Tetramethylrhodamine) or Fluorescein (FITC) were developed for that specific purpose [5].

6.1

Materials

Prepare all solutions using ultrapure water (prepared by purifying deionized water, to attain a sensitivity of 18 MΩ cm at 25
C) and analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise). Diligently follow all waste disposal regulations when disposing waste materials. 1. 1 PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2PO4, 1.8 mM KH2PO4 pH 7.4. Filter using a 0.22 μm membrane. 2. Fixing Solution: 4% of formaldehyde in PBS. 3. Wash Solution: 0.5% Triton-X-100. 4. Blocking Buffer: 5% normal serum, 0.5% Triton X-100 in PBS.

Tandem Ubiquitin Binding Entities (TUBEs)

6.2

Methods

199

All steps are performed at room temperature unless indicated otherwise. 1. Mount the prepared sample on 3-amino-propyl-tri-ethoxysilane (APES)-coated slides. 2. Fix the sample with the Fixing Solution for 20 min. 3. Wash four times with PBS for 10 min. 4. Wash the sample with Wash Solution for 10 min. 5. Apply 200 μL of Blocking buffer for 20 min. 6. Rinse the sample three times with PBS for 5 min each. 7. Add 2 μL of TUBE-TAMRA to 1 mL of Blocking Solution. Apply the TUBE-TAMRA solution to the sample for 1 h. 8. Wash the sample three times with PBS for 10 min. 9. Wash the sample with water for 2 min. 10. Cover the sample with 200 μL of antifade mounting reagent followed by a coverslip. 11. Image the sample using a fluorescence microscope at excitation wavelength of 540 nm and an emission wavelength of 578 nm.

6.3

Notes

1. Detection of polyubiquitination signal may require treatment of cells with a proteasome inhibitor such as MG-132 to permit the accumulation of polyubiquitinated proteins. 2. The use of TUBE-Fluorescein for imaging follows the same protocol as TUBE-TAMRA except for the imaging step, which requires excitation wavelength of 490 nm and an emission wavelength of 520 nm. 3. Imaging of polyubiquitinated proteins using linkage-selective TUBEs follows the same steps as those of the protocol corresponding to the fluorescent tag that they are conjugated to.

7 TUBEs for High-Throughput Cellular and In Vitro Assays to Monitor PROTAC® Driven Protein Ubiquitination and Degradation LifeSensors has engineered TUBEs to integrate in highthroughput screening platforms that can monitor and report the state of polyubiquitination on target proteins and their degradation profiles mediated by PROTAC® drug treatment. LifeSensors’ PROTAC® Assay Plate (PA950) is intended for the relative determination of the ubiquitination of a target protein in cellular lysates after PROTAC® treatment. While in vitro PROTAC assays are ubiquitination assays that are designed to monitor PROTAC® mediated ubiquitination on a TUBE-coated microtiter plate, these assays are designed to replace the more laborious,

200

Karteek Kadimisetty et al.

semiquantitative western blot methods to examine polyubiquitination and degradation of a target protein in cells and provide quantitative and reproducible results. Additionally, this assay allows for high-throughput screening to process compound libraries and establish rank order potency helping chemists establish SAR. The plate is designed for research use to accelerate the PROTAC® drug discovery process. Assay Principle: The PROTAC® Assay Plate is a sandwichbased assay in which polyubiquitinated proteins from cell lysates are captured in the wells of a precoated microtiter plate using a proprietary polyubiquitin binding reagent, TUBEs. Proteins that are not polyubiquitinated/unbound are removed by washing and then an antibody directed against the target protein is added followed by washing. Lastly, a secondary antibody conjugated to horse radish peroxidase (HRP) or a fluorescent label is used to measure the bound target antibody with detection reagents and a microplate reader. In vitro PROTAC® assays are high-throughput sandwich-based platforms designed to monitor ubiquitination of target recombinant protein when recruited by recombinant E3 ligase in presence of a PROTAC®, E1, E2, Ubiquitin, and ATP. In vitro PROTAC® assays are more relevant compared to simple proximity ligand assays since they report both ternary complex formation and ubiquitination, which is regulated by appropriate orientation of ternary complex that is mediated by a PROTAC®. Such in vitro ubiquitination assays will allow researchers to monitor ubiquitination profiles and identify potential degraders from a library of PROTAC®s with certainty to work in cell-based assays (Fig. 5a). 7.1

Materials

Prepare all solutions using ultrapure water (prepared by purifying deionized water, to attain a sensitivity of 18 MΩ cm at 25
C) and analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise). Diligently follow all waste disposal regulations when disposing waste materials. 1. 1 PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2PO4, 1.8 mM KH2PO4 pH 7.4. Filter using a 0.22 μm membrane. 2. PBS-T: 0.1% Tween-20 in 1 PBS. 3. Cell Lysis Buffer: 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 10% glycerol. Store at 4
C. Prior to use, add the following inhibitors to attain indicated final concentration: 50 μM PR-619, 5 mM 1,10-phenanthroline (o-PA), 2 mM PMSF, 25 μM MG132, Protease Inhibitor Cocktail.

7.2

Methods

1. See methods from Subheading 5 for cell-based PROTAC® assays. 2. See methods from Subheading 4 for in vitro PROTAC® assays.

Tandem Ubiquitin Binding Entities (TUBEs)

A

Excess Kinase Inhibitor + CRBN-Promiscious Kinase PROTAC

B dBET1

CRBN-Promiscious Kinase PROTAC 7

8.0 ¥10 6

UbMax CL Intensity

CL Intensity

6.0 ¥10

4.0 ¥10 7

2.0 ¥10

7

0 0.001

0.01

0.1

1

201

10

UbMax

6.0 ¥10 6 4.0 ¥10 6 2.0 ¥10 6 0 0.001

100

0.01

Log[Kinase PROTAC], µM

0.1

1

10

100

PROTAC, µM

C

dBET6

CL Intensity

4.0 ¥10

6

UbMax

3.0 ¥10 6 2.0 ¥10 6 1.0 ¥10 6 0 0.001

0.01

0.1

1

10

100

PROTAC, µM

Fig. 5 Validation of the TUBE-based platform for in vitro and cell-based PROTAC® molecules screening. (a) Ubiquitination profiles of recombinant Aurora Kinase A (AURKA) protein was demonstrated when treated with a Cereblon-based promiscuous kinase PROTAC® in an in vitro ubiquitination reaction. The target ubiquitination was monitored using anti-AURKA antibody in a TUBE-coated microtiter plate. The peak ubiquitination (UbMax) can be used to determine the potency of a target protein along with excess parent compound treatment in conjugation with PROTAC® treatment and will establish ternary complex formation. (b, c) Jurkat cells were treated with dBET1 and dBET6 PROTAC® (0.003–10 μM) for 45 min. Cell lysates (15 μg/well) were tested as described in the protocol and average chemiluminescence intensities were plotted against concentration of PROTAC  SD (n ¼ 3). The data suggests dBET6 is a more potent degrader than dBET1 with peak ubiquitination (UbMax) at 30 nM and 1 μM, respectively 7.3

Notes

Detection of polyubiquitination signal may require treatment of cells with a proteasome inhibitor such as MG-132 to permit the accumulation of polyubiquitinated proteins. To observe efficient polyubiquitination signal and subsequent degradation we recommend adhering to following guidelines. 1. To monitor polyubiquitination of target protein followed by its degradation, please select times before and after Dmax of PROTAC®. Also select dose-response studies based on predicted potency to demonstrate polyubiquitination and degradation profiles (Fig. 5b, c).

202

Karteek Kadimisetty et al.

2. Optimize antibody dilution and specificity of target ubiquitinated protein by western blotting. Selection of target antibody should be suitable for sandwich ELISA-based assays. 3. Optimize lysate concentration depending on target abundance. 4. Include appropriate controls like proteasome inhibitors and DUB inhibitors in conjugation with your PROTAC® treatment for best results and interpretation.

References 1. Meyer H-J, Rape M (2019) Enhanced protein degradation by branched ubiquitin chains. Cell 4:910–921 2. Xu P, Duong DM et al (2009) Quantitative proteomics reveals the function of unconventional ubiquitin chains in proteasomal degradation. Cell 1:133–145 3. Paiva S-L, Crews CM (2019) Targeted protein degradation: elements of PROTAC design. Curr Opin Chem Biol 50:111–119 4. Mattern M, Sutherland J et al (2019) Using ubiquitin binders to decipher the ubiquitin code. Trends Biochem Sci 44:599–615 5. Hjerpe R, Aillet F et al (2009) Efficient protection and isolation of ubiquitylated proteins using tandem ubiquitin-binding entities. EMBO Rep 10:1250–1258 6. Lopitz-Otsoa F, Rodriguez-Suarez E et al (2012) Integrative analysis of the ubiquitin proteome isolated using tandem ubiquitin binding entities (TUBEs). J Proteomics 75:2998–3014 7. Mikael A, Kramer HB et al (2011) Activitybased chemical proteomics accelerates inhibitor development for deubiquitylating enzymes. Chem Biol 18:1401–1412 8. Campbell MK, Sheng M (2018) USP8 deubiquitinates SHANK3 to control synapse density and SHANK3 activity-dependent protein levels. J Neurosci 38:5289–5301

9. Lear TB, Lockwood KC et al (2020) Kelch-like protein 42 is a profibrotic ubiquitin E3 ligase involved in systemic sclerosis. J Biol Chem 295 (13):4171–4180 10. Dybas JM, O’Leary CE et al (2019) Integrative proteomics reveals an increase in non-degradative ubiquitylation in activated CD4 T cells. Nat Immunol 20:747–755 11. Lonskaya I, Desforges NM et al (2013) Ubiquitination increases parkin activity to promote autophagic α-Synuclein clearance. PLoS One 8 (12):e83914. https://doi.org/10.1371/jour nal.pone.0083914 12. Loch CM, Strickler JE (2012) A microarray of ubiquitylated proteins for profiling deubiquitylase activity reveals the critical roles of both chain and substrate. Biochim Biophys Acta 11:2069–2078 13. Li Y, Hu Q et al (2019) PTEN-induced partial epithelial-mesenchymal transition drives diabetic kidney disease. J Clin Invest 129:1129–1151 14. Gillespie SR, Tedesco LJ et al (2017) The deubiquitylase USP10 regulates β1 and β5 and fibrotic wound healing. J Cell Sci 130:3481–3495 15. Im K, Mareninov S et al (2018) An introduction to performing immunofluorescence staining. Methods Mol Biol 1897:299–311. https://doi.org/10.1007/978-1-4939-89355_26

Chapter 11 Global Mass Spectrometry-Based Analysis of Protein Ubiquitination Using K-ε-GG Remnant Antibody Enrichment Alissa J. Nelson, Yiying Zhu, Jian Min Ren, and Matthew P. Stokes Abstract Ubiquitination is a post-translational modification that affects protein degradation as well as a variety of cellular processes. Methods that globally profile ubiquitination are powerful tools to better understand these processes. Here we describe an updated method for identification and quantification of thousands of sites of ubiquitination from cells, tissues, or other biological materials. The method involves cell lysis and digestion to peptides, immunoaffinity enrichment with an antibody recognizing di-glycine remnants left behind at ubiquitinated lysines, and liquid chromatography-tandem mass spectrometry analysis of the enriched peptides. Key words Mass spectrometry, LC-MS/MS, Proteomics, Ubiquitin, Ubiquitination, Protein degradation, Post-translational modification, PTMScan

1

Introduction Ubiquitin and ubiquitin-like modifiers such as NEDD8, ISG15, and SUMO1/2/3 are critical regulatory components of many cellular processes such as protein turnover, DNA replication and damage responses, normal growth and development, apoptosis, and a variety of diseases such as cancer [1–6]. Ubiquitin can be covalently linked to many cellular proteins by the ubiquitination process, which targets proteins for degradation by the 26S proteasome. Ubiquitin is first activated by forming a thioester complex with the activation component E1; the activated ubiquitin is subsequently transferred to the ubiquitin-carrier protein E2, then from E2 to ubiquitin ligase E3 for final delivery to the epsilon-NH2 of the target protein lysine residue [7–9]. Polyubiquitin chains can form using different residues on ubiquitin itself including K6, K11, K27, K29, K33, K48, K63, and the protein N terminus, and these diverse chains can have differing biological functions [10, 11]. Ubiquitin can be removed from polyubiquitin chains or from substrate

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021

203

204

Alissa J. Nelson et al.

proteins via the action of deubiquitinating enzymes (DUBs) [12, 13]. Development of proteomic tools to study ubiquitination on a global scale has greatly aided the ability to identify targets of specific ubiquitin ligases and deubiquitinases, as well as determine mechanism of action of compounds that affect the ubiquitination machinery. Among these advances was the development of antibodies that recognize the di-glycine remnant left at sites of ubiquitination after trypsin digestion through cleavage of the C-terminal –RGG sequence on ubiquitin (K-ε-GG) (Fig. 1) [14–16]. This Ubiquitin K-ε-GG remnant motif antibody will recognize the di-glycine remnant independent of flanking amino acid sequence (Fig. 2). Combination of antibody enrichment of K-ε-GG containing peptides with liquid chromatography-tandem mass spectrometry (PTMScan) [17] has allowed identification and quantification of thousands of sites of ubiquitination in a single experiment [15, 16, 18, 19]. In this method, cells, tissues, biofluids, or other biological materials are lysed and digested with trypsin, generating peptides as well as di-glycine remnants at sites of ubiquitination. Peptides are then reversed phase purified, dried, resuspended in immunoaffinity purification buffer, enriched with the K-e-GG antibody on beads, and run in LC-MS/MS. Quantification can be performed using labeling strategies such as SILAC or isobaric tagging [20, 21], or with label-free quantification using integrated peptide peak areas in the MS1 channel [22]. Previous protocols for PTMScan analysis have required large sample amounts (5–15 mg of protein per condition) and led to antibody release off beads during peptide elution, potentially providing a source of contamination for the analytical column on the instrument. The following improved, high sensitivity protocol (PTMScan HS) requires less material for the same number of identified K-ε-GG peptides and provides enriched peptides that are free of contaminating antibody, delivering cleaner samples for the mass spec analysis. The new PTMScan HS protocol also incorporates S-Trap digestion of peptides [23], a fast and efficient method to digest and clean up samples for subsequent processing that does not require a lyophilizer (as was needed in previous protocols).

2

Materials 1. 1
Phosphate-buffered Technology, 9808).

saline

(PBS)

(Cell

Signaling

2. Sodium Dodecyl Sulfate (SDS), reagent grade (Millipore Sigma L3771).

Profiling Ubiquitination with Mass Spectrometry

205

=Trypsin Cleavage Site

VLRLRGG

Trypsin

GG

K-ε-GG Antibody Ubiquitin

Substrate Protein

Fig. 1 Trypsin cleavage of ubiquitinated proteins generates peptides as well as K-ε-GG remnants at sites of ubiquitination. Rabbit monoclonal antibodies have been raised that specifically recognize that K-ε-GG remnant on any peptide irrespective of flanking amino acid sequence

Fig. 2 Sequence logo for K-ε-GG peptides identified using PTMScan antibody enrichment and LC-MS/MS analysis. There is no strong preference for any amino acid beyond the K-ε-GG residue. Logo generated with WebLogo (Stanford)

3. 1 M Triethylammonium Bicarbonate (TEAB), pH 8.0 (Thermo 90114). 4. Sonicator with microtip. 5. Polytron electronic homogenizer. 6. 2 mL microcentrifuge tubes and 15 mL conical tubes 7. Centrifuges capable of handling 15 mL tubes and 2 mL tubes. 8. Protein Quantitation Colorimetric Assay Kit, such as BCA (Thermo 23227). 9. Dithiothreitol (DTT) (Cell Signaling Technology 7016).

206

Alissa J. Nelson et al.

10. Iodoacetamide, PTMScan Qualified (Cell Signaling Technology 88931). 11. Phosphoric acid (Sigma 438081). 12. Methanol (Cell Signaling Technology 13604). 13. Water, Burdick & Jackson LCMS grade (CST 27732). 14. S-Trap Midi, 10 cartridges (Protifi C02-midi-10). 15. PTMScan Trypsin, TPCK-Treated (Cell Signaling Technology 56296). 16. 1 mM Hydrochloric acid (HCl). 17. Trifluoroacetic Acid (TFA), sequencing grade (Thermo 28904). 18. Acetonitrile (ACN), PTMScan Qualified (Cell Signaling Technology, 95031). 19. Vacuum concentrator (Speed-Vac). 20. PTMScan HS IAP Bind Buffer #1 (1
), contains 5% ACN (Cell Signaling Technology 25144). 21. PTMScan HS IAP Wash Buffer (1
) (Cell Signaling Technology 42424). 22. PTMScan HS KGG Remnant Magnetic Immunoaffinity Beads, 10 assays, 20 μL per assay (Cell Signaling Technology 34608). 23. Magnetic rack for 1.5 mL microcentrifuge tubes (Cell Signaling Technology, 14654). 24. End-over-end rotator. 25. Empore C18 disks (CDS Analytical 2215 or 2315). 26. 18 gauge blunt tip needle. 27. Pierce C18 Spin Tips (Thermo 84850). 28. Formic Acid, LCMS grade (Fisher A117-50 or similar quality for LCMS analysis). 29. C18 resin or C18 column such as Accucore C18 2.6 μm, 150 Å (Thermo). 2.1 Lysis and Digestion to Peptides

1. Dithiothreitol (DTT): Make 1.25 M stock: Resuspend one tube containing 192.8 mg with 1 mL water. Divide into 25 μL aliquots. Store at 20  C for up to 1 year. Thaw one aliquot for each experiment. 2. Iodoacetamide solution: Make 100 mM stock. Weigh out 19 mg of iodoacetamide and cover the tube with foil to protect it from light. Dissolve the powder in water to a final volume of 1 mL immediately before use. The iodoacetamide solution should be prepared fresh prior to each experiment. 3. 20% Trifluoroacetic acid: Add 1 mL of TFA to 4 mL of LCMS grade water. Store at room temperature.

Profiling Ubiquitination with Mass Spectrometry

207

4. 12% Phosphoric Acid: Mix 0.7 mL of 85% H3PO4 solution into 4.3 mL DI water. Store at room temperature. 5. 1
S-Trap Lysis Buffer: 5% SDS, 50 mM TEAB. Mix 250 mg SDS with 250 μL of 1 M TEAB and 4.75 mL water. Keep lysis buffer at room temperature to avoid precipitation of the SDS. 6. S-Trap Bind/Wash Buffer: 90% Methanol/100 mM TEAB. Mix 15 mL of 1 M TEAB pH 8.0 into 135 mL methanol. Store at room temperature tightly capped to avoid evaporation. 7. Digestion Buffer/S-Trap Elution A: 50 mM TEAB in LCMS grade water. Dilute 250 μL 1 M TEAB reagent to 5 mL with water. Store at room temperature. 8. S-Trap Elution B: 0.5% Trifluoroacetic acid (TFA) in LCMS grade water. Dilute 125 μL of 20% TFA with 4.875 mL LCMS grade water. Store at room temperature. 9. S-Trap Elution C: 50% Acetonitrile (ACN), 0.5% TFA in LCMS grade water. Mix 2.5 mL ACN, 2.375 mL water, and 125 μL of TFA. Store at room temperature. 2.2 Immunoaffinity Purification

1. IAP-Elution Buffer: 0.15% TFA in LCMS grade water. Mix 7.5 μL of 20% TFA into 1 mL of water. Prepare this solution in a container that has never been exposed to soap, as detergents will interfere with LCMS analysis.

2.3 C18 Sample Desalting

1. C18 tip Wetting Solution: 50% ACN, 0.1% TFA. Add 5 μL 20% TFA to 495 μL LCMS grade water, mix with 500 μL ACN. Store at room temperature in a well sealed tube/bottle to prevent evaporation of ACN. 2. C18 tip Equilibrate & Wash Solution: 0.1% TFA. Add 5 μL 20% TFA to 995 μL LCMS grade water. Store at room temperature. 3. C18 tip Elution Solution: 40% ACN, 0.1% TFA. Add 5 μL 20% TFA to 595 μL LCMS grade water, mix with 400 μL ACN. Store at room temperature in a well sealed tube/bottle to prevent evaporation of ACN.

2.4 LC-MS/MS Analysis

1. LCMS Solvent A: 3% acetonitrile in 0.12% formic acid. Add 1.2 mL formic acid to 968.8 mL LCMS grade water, mix with 30 mL ACN. Store at room temperature in a well sealed bottle to prevent evaporation of ACN. 2. LCMS Solvent B: 80% acetonitrile in 0.12% formic acid. Add 1.2 mL formic acid to 198.8 mL LCMS grade water, mix with 800 mL ACN. Store at room temperature in a well sealed bottle to prevent evaporation of ACN. 3. 50 cm
100 μm inner diameter capillary column packed with C18 reversed-phase resin.

208

3

Alissa J. Nelson et al.

Methods Carry out all steps at room temperature unless otherwise noted.

3.1 Lysis and Digestion to Peptides 3.1.1 Cell Line Protocol

1. Grow approximately 1–4
107 cells for each experimental condition (enough cells to produce approximately 1–2 mg of soluble protein). The cell number corresponds to approximately one 150 mm culture dish (depending on the cell type). 2. Wash adherent cells in cold PBS, then harvest by scraping in 500 μL of 1
S-Trap Lysis Buffer per plate. For suspension cells, collect the cells by centrifugation and wash the pellet with cold 1
PBS. Harvest the pellet with 500 μL of 1
S-Trap Lysis buffer. Do not cool lysate on ice as this may cause precipitation of the SDS (see Note 1). 3. Using a sonicator with a microtip, sonicate lysate at 5 W output with 3 bursts of 15 s each. Cool on ice for 1 min between each burst. Clear the lysate by centrifugation at 10,000
g for 15 min at room temperature (see Note 2) and transfer the protein extract (supernatant) into a new tube (see Note 3). Store the cleared lysate at 80  C at this point or continue directly through the protein quant and digestion steps.

3.1.2 Tissue Protocol

1. Select a fresh-frozen piece of tissue and place it in a 50 mL conical tube. Add approximately 2 mL of lysis buffer per 100 mg of wet tissue, or enough to submerge it completely. 2. Lyse the tissue using a homogenizer. Wash the tool with DI water in between each sample. Keep the moving parts submerged in buffer to avoid excessive foaming. 3. Using a sonicator with a microtip, sonicate lysate at 5 W output with 3 bursts of 15 s each. Cool on ice for 1 min between each burst. Clear the lysate by centrifugation at 10,000
g for 15 min at room temperature (see Note 2) and transfer the protein extract (supernatant) into a new tube (see Note 3). Store the cleared lysate at 80  C at this point or continue directly through the protein quant and digestion steps. 4. Measure soluble protein concentration using a colorimetric assay kit that is compatible with the SDS lysis buffer. The bicinchoninic acid (BCA) assay is recommended (see Note 4). 5. Normalize all samples so that equal amounts of protein are prepared for each condition and replicate. Alternatively, design the experiment so that equal numbers of cells are prepared for each sample. 6. Add 1/278 volume of 1.25 M DTT to the cleared cell supernatant to reach 4.5 mM DTT final concentration (e.g., 1.8 μL of 1.25 M DTT for 500 μL of protein extract). Mix well, and place the tube into a 55  C incubator for 30 min.

Profiling Ubiquitination with Mass Spectrometry

209

1. Cool the solution on ice briefly until it has reached room temperature. 2. Add 1/10 volume of iodoacetamide solution to the cleared cell supernatant, mix well, and incubate for 15 min at room temperature in the dark. 3. Add 1/10 sample volume of 12% phosphoric acid to the sample and mix well. 4. Transfer the sample to a 15 mL tube and add 6.6
the sample volume (e.g., 3.3 mL) of S-Trap Bind/Wash buffer (90% MeOH/100 mM TEAB). Mix well. A cloudy protein colloid should form immediately (see Note 5). 5. Load the full volume of sample onto the S-Trap Midi Cartridge, ensuring all the cloudy precipitate is transferred. Centrifuge 2 min at 4000
g at room temperature until all the solution has passed through. Discard the flow-through (see Note 6). 6. Wash the S-Trap with 3 mL of S-Trap Bind/Wash buffer (90% MeOH/100 mM TEAB) and centrifuge 2 min. Discard the flow-through. Repeat three times for four washes total. 7. Transfer cartridges to clean collection tubes prior to digestion. 8. Prepare 350 μL of trypsin working solution for each sample. Use trypsin at 1:10 (w:w) enzyme:substrate (e.g., 100 μg trypsin for 1 mg sample). Dilute 1 mg/mL trypsin stock up to 350 μL total with 50 mM TEAB and verify that the pH is ~8. Add 1 M TEAB, 1–2 μL at a time, to raise the pH if needed. The volume of digestion solution can be increased to ~500 μL if more trypsin stock is needed. 9. Add 350 μL trypsin solution to each cartridge and cap the tubes LOOSELY so that the digestion solution can absorb into the cartridge without creating negative air pressure. The cartridges may be centrifuged briefly but transfer any liquid that passes through back to the top of the cartridge. Place them in an incubator or water bath overnight at 37  C (see Notes 7 and 8). 10. The following day, add 500 μL of S-Trap Elution Buffer A (50 mM TEAB) to the S-Traps and centrifuge 1 min or until all solution passes through at 4000
g. Keep the cartridge in the same collection tube for all three elution steps. 11. Add 500 μL of S-Trap Elution Buffer B (0.5% TFA in water) to the S-Traps, centrifuge 1 min or until all the solution passes through, at 4000
g. 12. Add 500 μL of S-Trap Elution Buffer C (50% ACN/0.5% TFA in water) to the S-Traps, centrifuge 1 min.

210

Alissa J. Nelson et al.

13. Verify that the combined eluates have a pH < 3. If not, adjust the pH with a small amount of 20% TFA. Transfer the eluate (1.85 mL altogether) to a 2 mL microcentrifuge tube (see Note 9). 14. Dry the peptide solution in a vacuum concentrator (SpeedVac) set to ambient temperature overnight or until completely dry. The pellet should be visible at the end (see Notes 10–12). 3.2 Immunoaffinity Purification

1. Centrifuge the tube containing lyophilized peptide for 5 min at 2000
g at room temperature to collect all material for dilution in PTMScan HS IAP Bind Buffer #1 (see Note 13). Add 1.5 mL binding buffer to dried peptides. Resuspend pellets mechanically by pipetting repeatedly, taking care not to introduce excessive bubbles into the solution. Tubes can be shaken gently at room temperature on a vortexer or thermomixer for 5 min or placed in a sonicator bath for 2 min to ensure complete solubilization if necessary (see Note 14). 2. Clear solution by centrifugation for 5 min at 10,000
g at 4  C. There may be a small insoluble pellet. Cool on ice. Proceed to preparing the magnetic beads while waiting for this step. 3. Briefly spin the vial of antibody-bead slurry at no more than 2000
g for 2–5 s to bring down any buffer and beads clinging on the sides and cap of the vial. 4. Pipet antibody bead slurry gently to obtain a uniform suspension of beads, then take out 20 μL of bead slurry and place into a 1.7 mL microcentrifuge tube for each sample. Re-mix the bead stock before each pipetting step. Verify that each sample gets an equal aliquot of beads to ensure reproducible results. 5. Transfer 1 mL of ice-cold 1
PBS into the 1.7 mL microcentrifuge tube, mix buffer with beads by inverting tube five times. Place the tube on magnetic separation rack (12-Tube #14654 or 6-Tube #7017). Wait for 10 s or until beads are attracted to the magnet. Carefully remove PBS buffer. Repeat bead washing with 1 mL of 1
PBS three times. 6. Transfer the soluble peptide solution into the tube containing antibody beads and discard the insoluble pellet. Avoid creating bubbles upon pipetting. 7. Tighten the cap and seal the top of the tube with parafilm or caplock clips to avoid leakage. Incubate on an end-over-end rotator for 2 h at 4  C (see Note 15). 8. Briefly spin the tube at no more than 2000
g for 2–5 s to bring down beads and solution clinging to the sides and cap. Place the tube in the magnetic stand for 10 s. Transfer the peptide solution to a microcentrifuge tube. Optional: Save this

Profiling Ubiquitination with Mass Spectrometry

211

supernatant at 80  C for future use in subsequent IPs for other PTMs (see Note 16). 9. Add 1 mL of PTMScan HS IAP Wash Buffer to the beads, mix by inverting the tube five times. 10. Briefly centrifuge the tube, place the tube in the magnetic stand for 10 s and remove all Wash buffer. Repeat three more times for a total of FOUR Washes with PTMScan HS IAP Wash Buffer. 11. Add 1 mL cold LCMS grade water to the beads, mix by inverting tube five times. 12. Briefly centrifuge the tube, place the tube in the magnetic stand for 10 s and remove all water. Repeat one more time for a total of TWO water washes. 13. Add 50 μL of IAP Elution buffer (0.15%TFA) to the beads, use a mixer such as an Eppendorf Thermomixer set at 95%, and higher molecular weight bands near the top of the gel are observed, indicating successful polyubiquitination via all three methods. The ubiquitinated substrates are then purified by spin size exclusion (pur), which enriches for polyubiquitinated protein and removes free ubiquitin Table 3 Concentrations of components in ubiquitination reactions (see Note 25) Component

Rsp5

Keap1

Ubr1

10 buffer

1

1

1

Ovalbumin (mg/mL)

–

2

–

ATP (mM)

4

5

5

DTT (μM)

1

–

–

E1 (nM)

166

130

100

UbcH7 (μM)

2.9

–

–

UbcH5 (μM)

–

3

–

Ubc2 (μM)

–

–

10

Keap1 C151S (μM)

–

3

–

Rsp5 (μM)

2.9

–

–

Cul3/Rbx1 (μM)

–

3

–

Ubr1 (μM)

–

–

0.4

1.5

1.5

1.5

0.5

0.5

0.5

1.33

0.73

1.33

Substrate (μM) NADPH (mM)

a

Ubiquitin (mg/mL) a

Although NADPH is not a necessary component of the ubiquitination reaction, it is included to suppress ubiquitination of the single lysine remaining in DHFR

Proteasomal Unfolding Ability

235

3. Add ubiquitin and mix well to initiate the reaction. Save 1 μL of reaction mixture in 5 μL stop buffer at t ¼ 0 h for gel. 4. Incubate for 2 h at 30  C. Save 1 μL of reaction mixture in 5 μL stop buffer at t ¼ 2 h for gel. 3.3.2 Purification of Ubiquitinated Proteins

1. Load approximately 1.5 mL suspension of Sephadex G75 resin to empty mini-spin column. 2. Spin for 30 s at 600  g to remove ddH2O; empty collection tube. 3. Wash three times with 350 μL of PBS + BSA. Spin for 30 s at 600  g after each wash and empty the collection tube. 4. After third wash, spin one more time to remove excess PBS + BSA. 5. Place column in a clean 1.5 mL amber tube and add 60 μL of ubiquitination reaction to the resin. Spin for 30 s at 600  g to elute. 6. Add 60 μL of PBS + BSA to the resin and elute into the same amber tube by spinning. 7. Save 1 μL purified ubiquitinated substrate in 5 μL stop buffer for the gel. 8. Snap freeze substrate in liquid nitrogen and store at 80  C. 9. Assess the ubiquitination and purification by running an SDS-PAGE gel. Image with the Typhoon with Cy5 setting at 400 V. 10. Determine final concentration by quantifying the final purified ubiquitinated substrate relative to the ubiquitination reaction.

3.4 Proteasome Purification

Purification of proteasome, both from wild type (YYS40 [23]) and mutant strains, is accomplished by affinity chromatography through the addition of a 3X-FLAG-tag on the 19S regulatory particle’s Rpn11 subunit. We typically use low salt buffers and therefore purify the complete 26S proteasome with all associated proteins. Preparations with wild type proteasome often yield between 1.4 and 1.8 mL of proteasome with concentrations ranging from 300 to 500 nM. Mutant strains often yield lower concentrations than wild type. 1. Day 1: Start an overnight culture of the yeast strain, either from a plate or glycerol stock, in 5 mL of YEPDAU (or the appropriate selective media) at 30  C (see Note 26). 2. Day 2: Scale up the starter culture in a 4 L baffled flask containing 2 L of autoclaved YEPDAU. Shake for two nights at 30  C.

236

Christina M. Hurley and Daniel A. Kraut

3. Day 4: Spin down cells at 8224  g for 3 min or 4000  g for 10 min at 4  C. 4. Wash cells by resuspending with ice-cold ddH2O. Repeat spin from step 3. 5. Can store pelleted cells at 4  C overnight or store cells resuspended in Proteasome Buffer A00 (without adding 50 ARS or ATP) at 80  C. Otherwise proceed to step 6. 6. Resuspend cells in Proteasome Buffer A00 . 7. Lyse by two passages through an Avestin C5 high pressure homogenizer at 25,000 psi (see Note 21). 8. Check the pH of the lysate. If less than 7, raise to 7.5 with 1–2 mL 1 M Tris base. 9. Clarify by centrifugation two times at 48,384  g for 20 min at 4  C, transferring lysate to a new tube after the first spin. Filter supernatant with a 0.45 μm syringe filter (see Note 11). 10. Pre-wash 1 mL Anti-FLAG-agarose resin with Proteasome Buffer A0 . Resuspend resin with 1 mL Proteasome Buffer A0 and divide equally between two 50 mL falcon tubes (see Note 27). 11. Supplement the supernatant with 5 mM ATP and 1 ARS and add half to each tube. 12. Rotate tubes for 2 h at 4  C. 13. Pour incubated lysate and resin into empty PD-10 column and collect the resin. 14. Wash three times with 10 mL Proteasome Buffer A0 . 15. Spin column briefly (~5 s) at 500  g in swinging bucket rotor to remove excess buffer. 16. Incubate resin with one or less than one CV (1 mL) of Proteasome Elution Buffer for 15 min at room temperature. Typically, we use 0.75 mL of elution buffer for the first elution. 17. Recover by centrifugation as in step 15. Add to labeled 2 mL centrifuge tube and keep on ice. 18. Second elution typically uses 0.65 mL of Proteasome Elution Buffer. Repeat incubation and spin. Add second elution to 1.5 mL centrifuge tube and keep on ice. 19. Third elution typically uses 0.4 mL of Proteasome Elution Buffer. Repeat incubation and spin. Keep falcon tube on ice. 20. Use Coomassie dot blot to determine approximate proteasome concentration relative to known concentrations of BSA to identify which elutions to pool. 21. Determine proteasome concentration by Bradford method with BSA in Proteasome Buffer A0 as a standard.

Proteasomal Unfolding Ability

237

22. Aliquot, snap freeze in liquid nitrogen, and store at 80  C. 23. To assess the success of the purification can run proteasome on an SDS-PAGE gel to compare with old proteasome and/or use a native gel to assess activity. 24. Resin can be regenerated (see Note 24). 3.5 Unfolding Ability Assay

3.5.1 Assay

Our unfolding ability assay usually consists of 100 nM proteasome and 20 nM substrate, ensuring single turnover conditions. Multiple proteasome strains or substrate types can be tested in the same experiment. Reactions are started by adding proteasome to substrate. We typically run six to eight reactions per experiment (3–4 duplicate experiments), but as many as twelve reactions (6 duplicate experiments) can be run in one experiment using a multichannel pipette. Time courses may need to be adjusted based on the rate of degradation. The results of the assay are determined through SDS-PAGE, imaging on a Typhoon FLA 9500 or similar imager, and quantification of the amounts of full-length protein and fragment formed (Fig. 3). Curve-fitting and application of Eq. (1) give the unfolding ability of the proteasome. The following protocol is for a reaction with a single type of proteasome and three substrates, for a total of six 42 μL-reactions. Volumes are calculated based on stock concentrations of proteasome and substrates of 420 nM and 400 nM, respectively. The substrate master mix, proteasome mix, and individual substrate mixes have 10% more volume than needed to account for pipetting (Table 4). The final concentrations in the assay are: 50 mM Tris–Cl pH 7.5, 5 mM MgCl2, 5% glycerol (v/v), 1 mM ATP, 10 mM creatine phosphate, 0.1 mg/mL creatine phosphokinase, 2 mM DTT, 1% DMSO, 1 mg/mL BSA, and 500 μM NADPH. 1. Prepare a 96-well PCR plate with 9.3 μL of stop buffer in each reaction well at room temperature. 2. Set up reactions on ice. 3. Set up and label two sets of PCR strip tubes: one for proteasome and another for substrate. Also, label 0.6 mL tubes for proteasome mix, substrate master mix, Rsp5-substrate mix, Keap1-substrate mix, and Ubr1-substrate mix. 4. Make fresh 25 ARS by combining 4.62 μL 600 mM CP, 3.47 μL 8 mg/mL CPK, 2.77 μL 100 mM ATP, and 0.23 μL ddH2O. 5. Make the proteasome mix by adding 66.00 μL of proteasome, 5.54 μL 100 mM DTT, 129.4 μL ddH2O, and 23.56 μL 10 degradation assay buffer to the 25 ARS. Pipette up and down gently to mix (see Note 28). 6. Add 37.04 μL of the proteasome mix to each of the proteasome strip tubes.

238

Christina M. Hurley and Daniel A. Kraut

FL 10’’ 30’’

1’

2’

4’

8’

16’

32’

100

+

% of input

80 60 40 20 0 0

5

10

15 20 Time (min)

25

30

Fig. 3 Example of results from unfolding ability assays. A representative gel for a single experiment with wild type proteasome and Keap1-ubiquitinated substrate is shown on the left. Nonubiquitinated substrate (FL) serves as a size reference. The disappearance of ubiquitinated full-length substrate (boxed in black) and appearance of a DHFR-containing fragment (boxed in red) is observed. The full-length substrate (black, open squares) and DHFR fragment (red, open circles) intensities are quantified for each timepoint and plotted as percentages of the initial full-length substrate at the start of the reaction on the right. Curve fits are global fits to single exponentials. Error bars represent the SEM of 4 experiments

7. Make substrate master mix by adding 4.57 μL 10 degradation assay buffer, 10.16 μL 30 mg/mL BSA, 12.71 μL 12 mM NADPH (see Note 5), and 3.05 μL DMSO (see Note 29). Mix well. 8. Divide substrate master mix by adding 9.24 μL to each tube for the individual substrate mixes. Add 4.62 μL of substrate to its corresponding mix and mix well. Keep fluorescent substrates covered (see Note 30). 9. Place 6.30 μL of the appropriate substrate mixes in the substrate strip tube. 10. Briefly spin down the two sets of strip tubes and incubate for 5 min at 30  C. 11. During the incubation, prepare the multichannel pipettes. To start the reaction, a P50 or P200 multichannel works well. To take timepoints, a P10 or P20 multichannel works well to pull out 4 μL of reaction to add to the stop buffer. 12. Remove the lids of the strip tubes and with the multichannel pipette obtain 35.70 μL of proteasome mix. Check the tips to make sure that all tips are filled properly/uniformly. 13. Be ready to start the timer (counting up) as soon as proteasome mix is added to the substrate mix. Pipette up and down to mix thoroughly, but avoid creating bubbles. Good initial mixing is crucial to success of the assay. 14. Switch to the other multichannel pipette and take up 4 μL of reaction mixture to add to the first row of the 96-well plate. This should be completed at t ¼ 10 s. Pipette up and down to mix. Discard tips. If there is time before the next timepoint, put

6.30

1.75

0.42

12 mM NADPH (μL)

DMSO (μL)

Substrate mix 6.30 in reaction:

1.75

1.4

6.30

0.42

1.75

1.4

2.10

0.63

35.70

1.68 19.61 3.57

0.84 11.09 129.4 23.56

5.54

66.00

Proteasome

Add 6.30 μL of substrate mix per tube in substrate strip

0.92

3.85

Add 9.24 μL of substrate master mix to each substrate mix tube

3.05

12.71

10.16

–

4.62

3.08

4.57

Substrate master mixa

1.39

Substrate mixa

Add 37.04 μL of proteasome master mix per tube in proteasome strip

These mixes have 10% extra volume to account for pipetting errors

a

0.42

1.4

30 mg/mL BSA (μL)

2.10

2.10

400 nM substrate (μL)

35.70

35.70

0.63

1.68 19.61 3.57

1.68 19.61 3.57

0.63

0.84

0.84

10 buffer (μL)

420 nm Proteasome (μL) 100 mM DTT (μL) 25 ARS (μL) ddH2O (μL) 10 buffer (μL) Proteasome mix in reaction:

Rsp5Keap1Ubr1substrate substrate substrate mixa 10.00 10.00 10.00

Table 4 Example set-up for an unfolding ability assay

Proteasomal Unfolding Ability 239

240

Christina M. Hurley and Daniel A. Kraut

lids back on the substrate strip tube and the completed row of the 96-well plate. 15. Re-load tips for the multichannel pipette in preparation for the next timepoint. 16. Repeat the process for each timepoint until the end of the assay. For wild-type proteasome, a typical assay has timepoints at 0, 3000 , 10 , 20 , 40 , 80 , 160 , and 320 . In-between time points keep both assay and 96-well plate covered. Before each timepoint (time allowing) spin down the strip tube. 17. Once assay is complete, heat the 96-well plate at 95  C for 5 min to denature. Can store the plate at 20  C if necessary. Otherwise, proceed to running the gels. 3.5.2 Gels

We find that 9.25% Tris-Tricine gels separate full-length substrate and fragment bands well. Each gel should contain one complete time course per reaction, i.e., run one 8-well column from the plate. A nine-lane comb gives the right spacing for multichannel pipettes, which makes loading the gels much faster. We typically scan gels directly on a glass stage to reduce background fluorescence from the glass gel cassette: 1. Spread some ddH2O on the glass stage to make the gel easier to maneuver. 2. Carefully remove gel from the plates, trim the ends of the stacking gel, and place the gel on the wet glass stage. Make sure to remove bubbles before scanning. Scan at 350–400 V on a Typhoon FLA9500 (see Notes 31 and 32).

3.6

Data Analysis

Once the gel is imaged, we typically use ImageQuant to quantify the data; however, other software can be used (see Note 7). We quantify both the ubiquitinated/full-length and fragment bands for each timepoint. It is important to make sure quantification of the ubiquitinated/full-length substrate includes both ubiquitinated and nonubiquitinated substrate, which can be fairly spread out. This step ensures deubiquitination is not misinterpreted as degradation. Examples of the regions to quantify can be seen in Fig. 3. Next, transform the data so that the full-length and fragment data are percentages of the initial full-length substrate at the start of the reaction, with the full-length beginning at 100% and the fragment at 0% at the initial time point (which is set as t ¼ 0). We then graph the percent full-length and percent fragment versus time in a nonlinear curve-fitting program such as Igor Pro (Fig. 3), fitting the curves to single exponentials (see Note 33). The fits give amplitudes of full-length degradation and fragment formation. These amplitudes are then used in Eq. (1) to determine the unfolding ability. In the example from Fig. 3, the amplitude of

Proteasomal Unfolding Ability

241

full-length degradation is 70 and the amplitude of fragment formation is 13: U ¼ 70 13  1 ¼ 4.4.

4

Notes 1. We use Cy5, but other fluorophores can be used. 2. Perchloric acid is shock-sensitive. Do not use old reagents. Dispose of waste appropriately. 3. Any SDS-PAGE sample buffer should work. 4. Other Anti-FLAG resins we have tested do not work as well. 5. NADPH is often included in the reaction; however, it can be omitted. The result is an increase in the unfolding ability/ complete degradation of DHFR. 6. We use 9.25% Tris-Tricine. If using other types of gels when analyzing ubiquitinated proteins, be sure that all protein enters the running gel and none is trapped in the stacking gel. 7. We usually use ImageQuant for analysis because it has more sophisticated background subtraction; however, ImageJ is free. 8. Can start overnight cultures from a glycerol stock or from a colony on a plate. 9. After resuspending, can freeze cells indefinitely at 80  C. 10. Can use other cell lysis methods. 11. Pre-filtering the lysate with a 0.7 μm glass fiber filter prevents clogging of the 0.45 μm filter. 12. While we primarily purify proteins using an FPLC, we have also successfully purified proteins using gravity columns. 13. His-tagged SUMO protease is commercially available, or can make one’s own. 14. If nonspecific interactions prevent the protein from flowing over the column, the addition of 10% NPI-250 is normally sufficient to allow elution. 15. While we primarily purify labeled proteins on the FPLC, we have successfully purified them by gravity using Sephadex G75. 16. Sometimes nucleic acid contamination prevents concentration determination by A280. If so, use a Bradford assay to determine concentration. 17. Similar yields can also be obtained if grown in 2 L of LB media and induced at 0.5 mM IPTG overnight at 18  C. 18. C151S leads to elevated ubiquitination activity. 19. Keap1 is highly oxidizable.

242

Christina M. Hurley and Daniel A. Kraut

20. Aliquots can be diluted in water before measurement against an appropriate blank. 21. Can use other cell lysis methods; however, sonication does not work well. A third passage through the homogenizer may be necessary if the solution remains viscous or drops do not separate on second passage. We store the homogenizer at 4  C and cool the homogenizer with a heat exchanger and recirculating water bath at ~4  C while in use. 22. We typically use an empty PD-10 column. 23. We use a hole in the lid of a 50 mL conical tube. 24. Can strip the column with 100 mM Glycine pH 3.5, wash with PBS, and store in storage buffer at 20  C to reuse at least three times. Storage buffer: 1 PBS with 50% glycerol and 0.02% sodium azide. 25. There is a different 10 reaction buffer for each ubiquitination method. We store our ubiquitination enzyme stocks at 80  C and make many single-use aliquots from larger stocks to reduce freeze-thaw events. We keep the following stock solutions for nonenzyme components: 100 mM ATP, 30 μM DTT, 12 mM NADPH, and 20 mg/mL ovalbumin. 26. Overnight cultures tend to grow better when started from a plate. 27. Can complete this step during spins and directly filter into the falcon tubes with resin. 28. If using more than one type of proteasome, can make a master mix lacking proteasome and use that to make individual proteasome mixes. 29. 1% DMSO allows for the use of DMSO-soluble compounds, like proteasome inhibitors or methotrexate to stabilize DHFR. 30. A Styrofoam lid or some aluminum foil works well. 31. It is best to keep the stacking gel on, if possible, to avoid edge artifacts when quantifying ubiquitinated protein, which runs at the top of the gel. If the stacker comes off the resolving gel, add some water to the glass stage above the resolving gel. If it curls up, put more water on top. 32. If scanning through the glass plates, make sure that the glass is clean and free of gel debris; dry with a Kimwipe. Scan at 650 V. 33. Accumulation of fragment followed by reduction indicates trapping of some intermediate before it partitions. In this case, can fit to a double-exponential to determine overall amplitude, use kinetic modeling to determine U, or reduce the amount of proteasome to give single-exponential curves.

Proteasomal Unfolding Ability

243

Acknowledgments The authors thank Joseph Boscia IV, Mary Cundiff & Eden Reichard for comments and suggestions. This material is based upon work supported by the National Science Foundation under Grants No. 1515229 and 1935596 to DAK. References 1. Voges D, Zwickl P, Baumeister W (1999) The 26S proteasome: a molecular machine designed for controlled proteolysis. Annu Rev Biochem 68:1015–1068 2. Finley D, Ulrich HD, Sommer T et al (2012) The ubiquitin-proteasome system of Saccharomyces cerevisiae. Genetics 192:319–360 3. Finley D (2009) Recognition and processing of ubiquitin-protein conjugates by the proteasome. Annu Rev Biochem 78:477–513 4. Bard JAM, Goodall EA, Greene ER et al (2018) Structure and function of the 26S proteasome. Annu Rev Biochem 87:697–724 5. Lee C, Schwartz MP, Prakash S et al (2001) ATP-dependent proteases degrade their substrates by processively unraveling them from the degradation signal. Mol Cell 7:627–637 6. Prakash S, Tian L, Ratliff KS et al (2004) An unstructured initiation site is required for efficient proteasome-mediated degradation. Nat Struct Mol Biol 11:830–837 7. Groll M, Bochtler M, Brandstetter H et al (2005) Molecular machines for protein degradation. Chembiochem 6:222–256 8. Bar-Nun S, Glickman MH (2012) Proteasomal AAA-ATPases: structure and function. Biochim Biophys Acta 1823(1):67–82 9. Matyskiela ME, Martin A (2013) Design principles of a universal protein degradation machine. J Mol Biol 425:199–213 10. Finley D, Chen X, Walters KJ (2016) Gates, channels, and switches: elements of the proteasome machine. Trends Biochem Sci 41:77–93 11. Wehmer M, Rudack T, Beck F et al (2017) Structural insights into the functional cycle of the ATPase module of the 26S proteasome. Proc Natl Acad Sci U S A 114:1305–1310 12. Koodathingal P, Jaffe NE, Kraut DA et al (2009) ATP-dependent proteases differ substantially in their ability to unfold globular proteins. J Biol Chem 284:18674–18684 13. Kraut DA, Israeli E, Schrader EK et al (2012) Sequence- and species-dependence of proteasomal processivity. ACS Chem Biol 7:1444–1453

14. Reichard EL, Chirico GG, Dewey WJ et al (2016) Substrate ubiquitination controls the unfolding ability of the proteasome. J Biol Chem 291:18547–18561 15. Cundiff MD, Hurley CM, Wong JD et al (2019) Ubiquitin receptors are required for substrate-mediated activation of the proteasome’s unfolding ability. Sci Rep 9:14506 16. Vu N-D, Feng H, Bai Y (2004) The folding pathway of Barnase: the rate-limiting transition state and a hidden intermediate under native conditions. Biochemistry 43:3346–3356 17. Kim Y, Ho SO, Gassman NR et al (2008) Efficient site-specific labeling of proteins via cysteines. Bioconjug Chem 19:786–791 18. Raasi S, Pickart CM (2005) Ubiquitin chain synthesis. Methods Mol Biol 301:47–55 19. Carvalho AF, Pinto MP, Grou CP et al (2012) High-yield expression in Escherichia coli and purification of mouse ubiquitin-activating enzyme E1. Mol Biotechnol 51:254–261 20. Small E, Eggler A, Mesecar AD (2010) Development of an efficient E. coli expression and purification system for a catalytically active, human Cullin3-RINGBox1 protein complex and elucidation of its quaternary structure with Keap1. Biochem Biophys Res Commun 400:471–475 21. Eggler AL, Liu G, Pezzuto JM et al (2005) Modifying specific cysteines of the electrophile-sensing human Keap1 protein is insufficient to disrupt binding to the Nrf2 domain Neh2. Proc Natl Acad Sci U S A 102:10070–10075 22. Xia Z, Webster A, Du F et al (2008) Substratebinding sites of UBR1, the ubiquitin ligase of the N-end rule pathway. J Biol Chem 283:24011–24028 23. Saeki Y, Isono E, Toh-E A (2005) Preparation of ubiquitinated substrates by the PY motifinsertion method for monitoring 26S proteasome activity. Methods Enzymol 399:215–227 24. Kim HC, Huibregtse JM (2009) Polyubiquitination by HECT E3s and the determinants of chain type specificity. Mol Cell Biol 29:3307–3318

244

Christina M. Hurley and Daniel A. Kraut

25. Saeki Y, Kudo T, Sone T et al (2009) Lysine 63-linked polyubiquitin chain may serve as a targeting signal for the 26S proteasome. EMBO J 28:359–371 26. Yu H, Matouschek A (2017) Recognition of client proteins by the proteasome. Annu Rev Biophys 46:149–173 27. Martinez-Fonts K, Davis C, Tomita T et al (2020) The proteasome 19S cap and its ubiquitin receptors provide a versatile recognition platform for substrates. Nat Commun 11:477 28. Zhang DD, Lo S-C, Sun Z et al (2005) Ubiquitination of Keap1, a BTB-Kelch substrate adaptor protein for Cul3, targets Keap1 for

degradation by a proteasome-independent pathway. J Biol Chem 280:30091–30099 29. Kulathu Y, Komander D (2012) Atypical ubiquitylation - the unexplored world of polyubiquitin beyond Lys48 and Lys63 linkages. Nat Rev Mol Cell Biol 13:508–523 30. Choi WS, Jeong B-C, Joo YJ et al (2010) Structural basis for the recognition of N-end rule substrates by the UBR box of ubiquitin ligases. Nat Struct Mol Biol 17:1175–1181 31. Bodnar NO, Rapoport TA (2017) Molecular mechanism of substrate processing by the Cdc48 ATPase complex. Cell 169:722–735.e9

Part IV Methods for Evaluating Protein Degrader Function

Chapter 13 Methods for Quantitative Assessment of Protein Degradation Radosław P. Nowak, Hong Yue, Emily Y. Park, and Eric S. Fischer Abstract Assessment of small molecules that promote selective protein degradation (degraders) requires detailed characterization and measurement of protein levels in cells. Here we describe ratio-metric methods based on a dual fluorescent GFP/mCherry reporter system to quantify cellular protein levels. We further develop a kinetic framework for the analysis of such data. We describe two methods of generating the stable GFP-protein of interest (POI)/mCherry reporter cell lines, alternative readout methods by FACS and Laser Scanning Cytometry as well as the corresponding tools used for processing and analysis of such data. Finally, we show that the commonly used half-maximal degradation constant (DC50) or maximum degradation efficacy (Dmax) metrics are time-dependent and propose a time-invariant Michaelis-Menten-like analysis of degradation kinetics with analogous key parameters Km app and Vmax app. Key words Degradation kinetics, Targeted protein degradation, PROTAC, IMiD, DC50, CRBN, VHL

1

Introduction Small molecule-induced protein degradation is a rapidly growing field and quantitative monitoring of protein levels upon degrader treatment is a key experiment validating its design [1–3]. Proteasomal degradation is a multi-step process that involves ATP-dependent activation of ubiquitin by an ubiquitin-activating enzyme (E1), conjugation of ubiquitin (Ub) to a ubiquitinconjugating enzyme (E2), E2-Ub binding to the E3-ligase engaged with substrate, ubiquitination of the substrate, ubiquitin chain elongation, shuttling to and processing by the proteasome leading to proteasomal degradation [4]. Under the assumption that a single degrader molecule is able to be recycled and promote ubiquitination of another substrate, small molecule degraders can be viewed as catalysts of this enzymatic reaction, allowing for dosedependent recruitment of the substrate to the E3 ligase [5, 6]. Their action on the protein levels resembles classical

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021

247

Radosław P. Nowak et al.

A

B

GFP/mCherry ratio

dBET6 Time series 1.0

1.00E-05 3.33E-06 1.11E-06 3.70E-07 1.23E-07 4.12E-08 1.37E-08 0

0.5

MZ1 Time series

GFP/mCherry ratio

248

0.0

1.0

1.00E-05 2.00E-06 4.00E-07 8.00E-08 1.60E-08 3.20E-09 6.40E-10 0

0.5

0.0 0

2

4

6

8

0

2

Time (h)

C

D

6

8

MZ1 Km app

2.0

2.0

1.5

1.5 Velocity

Velocity

dBET6 Km app

1.0 0.5

1.0 0.5

0.0 0

0.0

2×10 -6 4×10 -6 6×10 -6 8×10 -6 1×10 -5

0

2×10 -6 4×10 -6 6×10 -6 8×10 -6 1×10 -5

Concentration [M]

E

Concentration [M]

F

dBET6 DC50

MZ1 DC50

10-6

10 -6

10-7

10 -7

DC 50 [M]

DC 50 [M]

4 Time (h)

10-8

10 -8 10 -9

10-9

10 -10

10-10 0

2

4

6

Time (h)

8

0

2

4

6

8

Time (h)

Fig. 1 (a) Degradation time series for BRD4BD2 GFP/mCherry cell line [8]. Note the hook-effect visible at 10 μM concentration of dBET6, CRBN-based BRD4 degrader. Data is fitted to one phase decay function in GraphPad Prism 7.0. Each data point is from at least 2000 events, n ¼ 1. (b) Same as in (a) but for MZ-1, VHL-based BRD4 degrader. No hook effect is observed at 10 μM concentration. (c) Michelis-Menten-like characterisation of CRBN based BRD4 degrader dBET6. Initial velocity estimated from one phase decay is plotted against degrader concentration in GraphPad Prism 7.0. 10 μM concentration point is omitted due to hook effect. Km app and Vmax app calculated by fitting Michelis-Menten equation assuming total enzyme concentration (Et) of 1. (d) Same as in (b) but for VHL based BRD4 degrader MZ-1. (e) Plot of DC50 dependence on the incubation time for dBET6. (f) Same as in (c) but for MZ-1. All data in Fig.1 were analyzed by flow cytometry

Michaelis-Menten kinetics (Fig. 1c, d), where local substrate concentration tracks the levels of small molecule degrader. As any enzymatic process, the extent of degradation is dependent on

Methods for Quantitative Assessment of Protein Degradation

249

protein concentrations of individual components, catalytic efficiency of the enzymes, and for small molecule driven degradation, its concentration and time of exposure (Figs. 1a, b and 2). Since concentrations of the proteins are determined by the cellular context, a frequent representation of the degrader efficacy is a degrader dose-response curve plotting the degrader concentration versus target protein level. This representation allows for the calculation of the DC50 (concentration where response reaches 50% of the maximum) and Dmax (the maximum degradation efficacy). It is important to note that both of these parameters are timedependent (Figs. 1e, f and 2), and cannot be directly compared between different treatment times and cell lines, unless performed at steady state, which needs to be determined experimentally. When referring to DC50 and Dmax, we therefore propose to use the respective incubation time in the parameter name, as an example for 5 h incubation, we would measure DC50, 5h and Dmax, 5h. The most widely used method to monitor protein levels, and hence the efficacy of degraders, is western blotting. While these can be performed with relative ease when antibodies are available, the quantification of protein levels is at times difficult, prone to loading artifacts and limits the comparison between compounds to major changes in band intensities. Alternative quantitative methods that can be adapted to 96- and 384-well plate formats include monitoring of protein levels with fluorescent protein fusions, luciferase fusions or a recently developed split luciferase system, where HiBit peptide-tagged protein fuses with LargeBit forming a functional NanoLuc luciferase allowing for accurate protein level quantification [7]. In this chapter, we describe quantitative ratio-metric methods used to assess protein degradation kinetics focusing on cellular systems using fluorescent reporters and further provide methods of analysis of such data. In this cellular assay, a protein of interest is expressed as GFP fusion followed by a P2A cleavage site or internal ribosomal entry site (IRES) and mCherry, which results in expression of GFP-POI fusion and mCherry coupled by virtue of being expressed from a single mRNA transcript (Fig. 3). Upon treatment with a degrader molecule, the GFP-fusion is degraded but the mCherry is unaffected allowing for a ratio-metric control of protein levels by measuring the GFP/mCherry ratio (Fig. 4). We demonstrate the application of this technique by monitoring kinetics of degradation for CRBN binding IMiD degraders (lenalidomide) (Fig. 2) as well as CRBN (dBET6) or VHL recruiting (MZ-1) BRD4 PROTACs [8, 9] (Fig. 1 and Table 1). This robust assay allows for estimation of time-invariant parameters Kmapp and Vmaxapp by fitting in a Michelis-Menten like-model (Eq. 1).

250

Radosław P. Nowak et al.

A

Lenalidomide Time series

B

Lenalidomide Km app 10

2.00E-06

1.0

8

4.00E-07 Velocity

GFP/mCherry ratio

1.00E-05

8.00E-08 1.60E-08

0.5

6 4

3.20E-09

0.0 0

2

4

6

6.40E-10

2

0

0 0.0 5.0 10-7 1.0 10-6 1.5 10-6 2.0 10-6 2.5 10-6

8

Time (h)

Concentration [M]

C

Lenalidomide DC50

10-6

DC 50 [M]

10-7 10-8 10-9 10-10 0

2

4

6

8

Time (h)

Fig. 2 (a) Degradation time series for IKZF1Δ1–82,Δ197–239,Δ256–519-GFP/mCherry reporter cell line [8]. Data is fitted to one phase decay function in GraphPad Prism 7.0. Each data point is from at least 2000 events, n ¼ 1. (b) Michelis-Menten-like characterisation of lenalidomide degradation of IKZF1 reporter. Initial velocity estimated from one phase decay is plotted against degrader concentration in GraphPad Prism 7.0. Kmapp and Vmax app calculated by fitting Michelis-Menten equation assuming total enzyme concentration (Et) of 1. (c) Plot of DC50 dependence on incubation time for lenalidomide

A GFP

POI

p2a/IRES

mCherry

B POI O N

O NH

NH2

O

degrader

GFP

mCherry

mCherry

Fig. 3 (a) Map of the GFP-POI/mCherry reporter construct. (b) Cartoon diagram of protein degradation in the reporter system. Treatment with a degrader results in removal of GFP-fusion protein leaving mCherry levels unaffected

Methods for Quantitative Assessment of Protein Degradation

104

10 µM Len, 6.25h DMSO, 6.25h

103

102

Gate 1 87.3% 101

100 100

101

10 2

103

GRGreen Fluorescence (GRN-HLog)

B

A

Side Scatter (SSC-HLog)

251

104

103

10 µM Len, 6.25h DMSO, 6.25h

102

101

100

104

100

101

10 2

103

104

Red Fluorescence (RED-HLog)

Forward Scatter (FSC-HLog)

Fig. 4 (a) Side scatter (SSC) and forward scatter (FSC) plot of IKZF1-GFP/mCherry reporter cell line treated with DMSO or 10 μM lenalidomide for 6 h 20 min. Data analyzed on Guava flow analyzer; 3000 events are shown. Suggested gate for FSC and SSC is highlighted with black oval. (b) Data from gate 1 in (a) plotted as GFP and mCherry fluorescence signal Table 1 Calculated Km

app

and Vmax

app

corresponding to Figs. 1c, d and 2b.

dBET6

MZ-1

Lenalidomide

Vmax app

1.06  0.06 A.U.

1.46  0.13 A.U.

7.35  0.45A.U.

Km app

429  82 nM

23  10 nM

77  21 nM

Y ¼ V max app

X , where V max app ¼ E t K m app þ X

app kcat app

ð1Þ

where X is the degrader concentration, and Y is the corresponding velocity extracted from time-course of degradation and Et app represents the enzymatic components of the protein degradation pathway with a single enzyme concentration. Parameters in this model approximate the proteasomal degradation process as a single enzymatic reaction, which is dominated by the rate limiting components of the system. For example, if the E3 ligase processing is the rate limiting step in degradation process, the Et app is dominated by concentration of the E3 ligase, [CRL4CRBN], similar for Km app and Vmax app. While Km app and Vmax app can be directly calculated and compared between degraders, the calculation of the related catalytic turnover parameter kcat app requires estimation of the total enzyme concentration Et app which requires detailed characterization of the degradation system.

252

Radosław P. Nowak et al.

The classical Michelis-Menten model describes the relationship between substrate concentration and velocity. Eq. (1) assumes a linear relationship between substrate and degrader concentrations, which breaks down at high compound concentrations for both molecular glues or PROTAC molecules. Since titration of a degrader allows for recruitment of the substrate to the E3 ligase, local concentration of the substrate for a given degrader concentration follows occupancy driven model described by sigmoidal function for IMiD like degraders or a ternary complex model for PROTACs [10], both of which display non-linear behaviour. The detailed kinetic degradation characteristics of molecular glue and PROTAC systems also depends on multiple other factors such as E3 ligase concentration, total substrate concentration, and their respective affinities to the degrader molecules. Despite the limitations of the model, both kinetic parameters that it approximates have a familiar interpretation. Km app represents concentration of the substrate (target protein) needed to reach half maximum velocity, a measure of substrate affinity, Vmax app represents maximum degradation rate achieved by the system at saturating substrate concentration, and kcat app represents the turnover number of the enzyme specifying the maximum number of substrate molecules degraded per unit of time. Protein degradation is a multistep process and catalytic parameters estimated here are an approximation of the whole system, rather than a specific individual component. In an example presented here, MZ-1 has lower Km app value (23  10 nM) than dBET6 (429  82 nM) indicating tighter affinity to the substrate, with similar maximal degradation rates between the two E3 ligase systems (Table 1). While protein expression levels and cellular localization have to be taken into account when comparing degraders based on different E3 ligase complexes, characterization of degradation kinetics with quantitative methods presented here allows for informed design of degraders and provides time-invariant parameters such as Km app and Vmax app that can be used in iterative optimization cycles.

2

Materials

2.1 Preparation of a Reporter Plasmid for Flp293T

1. Flp293T cells (Thermo Fischer). 2. pOG44 plasmid (Invitrogen). 3. FlipIn plasmid with GFP-POI/mCherry (Fischer Lab). 4. Opti-MEM media. 5. DMEM media with 10% FBS. 6. DMEM media with 20% FBS and 10% DMSO. 7. Lipofectamine 2000.

Methods for Quantitative Assessment of Protein Degradation

253

8. FuGene HD. 9. Hygromycin-B. 10. 6-well plate, tissue culture-treated. 11. 10 cm petri dish plate, tissue culture treated. 12. 1.5 mL Eppendorf tubes. 13. 37  C Mammalian Tissue Culture Incubator. 2.2 Preparation of Reporter Line by Lentivirus

1. HEK 293T cells. 2. Opti-MEM media. 3. DMEM with 10% FBS. 4. FuGene HD. 5. 6-well plate, tissue culture-treated. 6. 10-cm petri dish plate, tissue culture-treated. 7. 1.5 mL Eppendorf tubes. 8. Reporter constructs. Suitable lentiviral vectors include C-terminal GFP fusion/mCherry reporters Artichoke (Addgene plasmid # 73320; http://n2t.net/addgene:73320; RRID:Addgene_73320) or Cilantro 2 (Addgene plasmid # 74450; http://n2t.net/addgene:74450; RRID: Addgene_74450). 9. psPax2 plasmid available at Addgene (plasmid # 12260; http:// n2t.net/addgene:12260; RRID:Addgene_12260). 10. pVSV-G plasmid available at Addgene (plasmid # 36399; http://n2t.net/addgene:36399; RRID:Addgene_36399). 11. Polybrene. 12. Puromycin. 13. Liquid nitrogen. 14. Benchtop centrifuge. 15. 37  C Mammalian Tissue Culture Incubator.

2.3 Flow Cytometry Analysis

1. Cell Analyzer with green and red lasers (for example, Guava easyCyte HT, Millipore). 2. FlowJo (LCC) or similar data analysis program. 3. GFP-POI/mCherry stable cell line. 4. 96-well, tissue culture-treated plate. 5. 96-well, round bottom plate. 6. 8- or 12-channel pipette or multistepper pipette. 7. DMEM with 10% FBS. 8. Trypsin 0.25%.

254

Radosław P. Nowak et al.

9. Compounds and DMSO solution. 10. 37  C Mammalian Tissue Culture Incubator. 2.4 Laser Scanning Cytometry Analysis

1. 384-well black plate with clear flat bottom 2. FluoroBrite DMEM media with 10% FBS. 3. Multichannel pipette p125 or similar. 4. Compound dispenser (for example, D300e Digital Dispenser (HP)). 5. 37  C Mammalian Tissue Culture Incubator. 6. Laser Scanning Cytometer (Acumen, TTP Labtech).

3

Methods

3.1 Preparation of Reporter Cell Line 3.1.1 Preparation of a Reporter Cell Line in Flp-in System

1. The evening before transfection, prepare a 6-well plate with Flp293T cells at density of 0.5  106 cells/well in 1.5 mL 10% FBS-containing DMEM media (see Note 1). 2. Day of transfection. Replace media with 1.9 mL serum-free OptiMEM. 3. Prepare a DNA/pOG44/transfection reagent mixture in serum-free Opti-MEM (see Note 2). (a) Separately mix A: 0.2 μg plasmid, 1.8 μg of pOG44, and Opti-MEM in 50 μL volume in Eppendorf tube. Incubate for 5 min. (b) Separately mix B: 5 μL of Lipofectamine with Opti-MEM 45 μL. Incubate for 5 min (see Note 3). (c) Combine mix A and mix B into Eppendorf tube to create mix C. Incubate 10–15 min. 4. Add mix C drop-wise to wells. Swirl plate gently. Place in a mammalian tissue culture incubator. 5. After 6–8 h, replace with 10% FBS-containing DMEM media (see Note 4). 6. When confluent, trypsinize and transfer the cells to a 10 cm petri dish plate with 50 μg/mL Hygromycin B, top up with 10% FBS-containing DMEM media to total volume of 10 mL (see Note 5). 7. Maintain cells under Hygromycin B selection, changing to media with fresh antibiotic every 2 days (see Note 6). 8. After 1 week under selection, remove hygromycin B from growth medium let grow to high confluency (see Note 7). 9. Expand your new cell line in DMEM media containing 10% FBS and freeze it in DMEM media containing 20% FBS and 10% DMSO.

Methods for Quantitative Assessment of Protein Degradation 3.1.2 Preparation of a Reporter Line by Lentivirus Transduction Lentiviral Packaging

255

1. On the day before transfection, seed 0.7  106 HEK 293T cells per well in a 6-well plate in DMEM media containing 10% FBS. The final volume is 3 mL. 2. Next day in the afternoon, cells should be 70–80% confluent. 3. Co-transfect with reporter construct and lentiviral packaging vectors. Mix 1 μg construct DNA with 1 μg psPax2 and 0.1 μg pVSV-G. Dilute DNA mixture with Opti-MEM up to 50 μL. 4. Dilute FuGene HD. Mix 8 μL FuGene HD in 42 μL OptiMEM. 5. Combine the diluted DNA mixture with diluted FuGene HD. Add the diluted FuGene HD to the diluted DNA mixture and mix gently by pipetting up and down for five times. 6. Allow transfection complexes to form for 15 min at room temperature. Add transfection complexes drop-wise to the cells and mix gently by rocking plate back-and-forth and sideto-side (see Note 8). 7. The next morning, aspirate media and replace with 2 mL of DMEM media containing 10% FBS. 8. At approximately 48 h after transfection, collect the supernatant (your lentivirus) and centrifuge for 5 min at 1000  g. You can check the cells for fluorescence at this point to monitor transfection efficiency (see Note 9). 9. Proceed to lentiviral transduction or aliquot the lentivirus into 1.5 mL Eppendorf tubes, flash freeze with liquid nitrogen, and store at 80  C until use.

Lentiviral Transduction

1. Seed cell line of interest at 0.55  106 cells per well in a 6-well plate. The final volume is 2 mL. 2. The next afternoon, prepare transductions. Remove 1 mL media from wells (1 mL remaining), work with virus harvested on the same day or quickly thaw frozen viruses at 37  C. Mix virus with polybrene in DMEM media containing 10% FBS. The final concentration of polybrene is 8 μg/mL (see Note 10). 3. Add the virus/polybrene/DMEM mixture to the cells dropwise. 4. Next morning, replace the media on the cells with 2 ml fresh DMEM containing 10% FBS. 5. Check cell each day to ensure cells are not confluent. Expand to 10 cm petri dish plate when cells are confluent (follow the regular mammalian cell culture protocol for splitting using BSL2 lentivirus safety). 6. At 72 h after transduction, add appropriate antibiotic for selection at a concentration based on prior kill curve for your specified cell line (see Note 11).

256

Radosław P. Nowak et al.

7. Monitor the cells. During selection, continue to replace media with fresh antibiotic-containing media every 2 days. 8. Selection is typically complete after 1 week. Lentivirustransduced cells should grow at a normal rate under selection, while 100% of nontransduced cells should die at the same antibiotic concentration. 9. After selection, positive reporter cells could be accumulated further by sorting for GFP and mCherry double positive cells with Fluorescence-activated cell sorting (FACS). 3.2 Degradation Readout 3.2.1 Flow Cytometry Analysis Cell Seeding and Treatment

1. Seed 30,000 cells of reporter cell line in each well of 96-well plate the day before treatment. Use 100 μL final volume of DMEM containing 10% FBS media. Ideally, cells are ~50% confluent at the time of treatment (see Note 12). 2. Next day, treat cells with compounds and DMSO control, note down the treatment time. Final percentage of DMSO in the well should be normalized to 0.5% or lower (see Note 13). 3. To take a time point, aspirate media from the plate (see Note 14). 4. Add 25 μL of trypsin to each well with a multistepper or multichannel pipette and incubate at 37  C for 5 min. Use light microscope to ensure cells are detached. 5. Add 100 μL of DMEM containing 10% FBS media to trypsinized cells (see Note 15). Resuspend the cell suspension by pipetting up and down at least three times through a p200 tip to generate single cell suspension (see Note 16). 6. Transfer the entire cell suspension to a 96-well round bottom plate compatible with Guava Cell Analyzer (Guava easyCyte HT).

FACS Analyzer Readout

The method presented here in brief is used for Guava Cell Analyzer (Guava easyCyte HT), but principles are translatable to other cell analyzers. 1. Make a worklist in the Worklist Editor for Guava to acquire at least 2000 events per well (see Note 17). 2. Start the Guava software. Start the worklist (created in step 1). Load plate into the Guava. 3. Adjust the signal (the forward and side scatter and the GFP and mCherry channels). The easiest way is to change the GFP and RFP dials such that the GFP/mCherry plot appears as a line with a slope of 1 passing through the origin (see Note 18). 4. Once the signal has been adjusted, start the run (see Note 19).

Methods for Quantitative Assessment of Protein Degradation Data Analysis

257

The data analysis can be completed with any compatible FACS analysis program. FlowJo is used below as an example. 1. Import FCS files into FlowJo. 2. Edit columns and add “file name” column to identify samples. 3. Double click a DMSO control sample (click the file name). 4. Select cells based on forward and side scatter (see Fig. 2a). (a) Change the y-axis to side scatter (log) and the x-axis to forward scatter (log). (b) Click the oval button at the top of the window to make an oval selection around the densest forward scattered cells (generally the highest intensity) (see Note 20). 5. Apply analysis to all wells. (a) Return to main window, and “copy analysis” to the group. 6. Select cells based on GFP and mCherry (see Fig. 2b). (a) Return to the plot window and change the y-axis to green fluorescence (log) and the x-axis to red fluorescence (log). (b) Click the rectangle button at the top of the window and select an area around the FACS sorted cells that captures all of the GFP cells but removes cells with the lowest intensity mCherry signal. 7. Apply analysis to all wells. (a) Return to main window, and “copy analysis” to the group. 8. Calculate 10 * GFP/mCherry ratio for each event. (a) With the file name of the DMSO control selected, go to the “Tools” tab in the main window and select “Derive parameters”. (b) In the formula box, type “10*green fluorescence (log)/ red fluorescence (log)” (see Note 21). (c) Green and red fluorescence variables can be picked from the “insert reference” tab in the window. 9. Apply analysis to all wells. (a) In main window, “copy analysis to group” for the derived values. 10. Generate median of the GFP/mCherry ratio for each well. (a) Select the gated GFP/mCherry cells box, and in the Workspace tab, choose “Median” and select “Derived”. 11. Apply analysis to all wells. (a) Return to main window, and “copy analysis” to the group. 12. Save the file and export a version to Excel for making a graph. 13. Spot check a few samples by overlaying gates (see Note 22).

258

Radosław P. Nowak et al.

3.2.2 Laser Scanning Cytometry Analysis

Cell Seeding and Treatment

The ratio-metric fluorescence signal can be read and analyzed by any plate-based (confocal) microscope or laser cytometer reader capable to extract GFP and mCherry intensities. The laser scanning cytometry is used as a readout with the Acumen High Content Imager (TTP Labtech) using 488 nm and 561 nm lasers as an example instrument. Compound dispensing can be done on alternative systems, d300 (HP) is given as an example. 1. Seed reporter cells on 384-well plate (black, see through bottom, tissue culture-treated). Seeding density: 30% of a confluent 10 cm petri dish plate is used for a single 384-well plate. (a) Harvest cells by centrifuge at 1000  g speed and then resuspend cell pellets in 22 mL of Fluorobrite DMEM media (with 10% FBS). 2. Mix well and dispense 50 μL cells each well by multichannel pipette. 3. Spin the plates 500  g for 1 min. 4. Put cells back into incubator and incubate at 37  C for overnight. 5. Treat the cells in the morning using the HP D300e Digital Dispenser. (a) Dispense the compounds of interest and controls at the same time and normalize the DMSO amount per well to 0.5% (see Note 23). 6. Incubate at 37  C for the respective time before imaging. 7. Read the plates by Laser Scanning Cytometer.

Imaging Readout

1. Clean the bottom of 384-well plate by 75% EtOH and dry with Kimwipe. 2. Put the plate into the Laser Scanning Cytometer (Bottomreading fluorescent imager). 3. Scan each well by Green fluorescent signal (Excitation laser: 488 nm; Filter: 500–530 nm) and Red florescent signal (Excitation laser: 561 nm; Filter: 575–640 nm) individually. 4. The image files are saved and exported for analysis.

Data Analysis

Data analysis can be performed using the inbuilt data analysis software, typically provided with the instrument. The example below is using Cell Profiler and is optimized to HEK293T cells. Latest version of Cell Profiler can be downloaded from https:// cellprofiler.org/ [11]. The general principle of the analysis is to identify GFP positive, mCherry positive cells

Methods for Quantitative Assessment of Protein Degradation

259

1. Install Cell Profiler. 2. Create a analysis pipeline file including following parts and settings (default settings unless otherwise stated): (a) Images. l

Import Acumen images.

(b) Meta Data. l Extract Meta Data regular expression: (?P.*). *.(?P[0-9]{3}).(?P[A-P][0-9] {1,2}).* (c) NamesAnd Types. l

GFP—488 channel.

l

RFP—561 channel.

(d) Groups. (e) Align. l

Mutual Information-based.

l

Input GFP!output Aligned Red.

l

Input RFP!output Aligned Green.

(f) Crop. l

Input Aligned Red!output Crop Aligned Red.

l

Crop center coordinates x: 2000 y: 1000, eclipse radius x: 1000, y: 500.

(g) Crop. l

Input Aligned Green!output Crop Aligned Green.

l

Crop center coordinates x: 2000 y: 1000, eclipse radius x: 1000, y: 500.

(h) Correct Illumination Calculate (create background function for Red channel). l

Input Crop Aligned Red!output Illum Aligned Red.

l

Box size 60.

l

Gaussian Smoothing, automatic filter size.

(i) Correct Illumination Calculate (create background function for Green channel). l

Input Crop Aligned Green!output Illum Aligned Green.

l

Box size 60.

l

Gaussian Smoothing, automatic filter size.

(j) Correct Illumination Apply (subtract the background). l

Input Crop Aligned Red, Illum Aligned Red!output Corr Aligned Red.

l

Input Crop Aligned Green, Illum Aligned Green!output Corr Aligned Green.

260

Radosław P. Nowak et al.

(k) Enhance Or Supress Features (see Note 24). l

Input Crop Aligned Green!output Filtered GFP.

l

Speckles size 30.

(l) Identify Primary Objects (find red cells). l

Input Corr Aligned Red!output Red Cells.

l

Dimeter 8, 60 pixels.

l

Global Threshold by Robust Background method.

l

Lower outlier fraction 0.05.

l

Upper outlier fraction 0.05.

l

Mean averaging method, 2 standard deviations, Thresholding smoothing scale 1.3488, correction factor 0.9, lower bound 0, upper bound 1.

(m) Threshold (Threshold the speckle-enhanced green channel to distinguish GFP-positive from GFP-negative areas). l

Input Filtered GFP!output Thresh GFP Well.

l

Robust background method, parameters as in l.

(n) Mask Objects (Call a cell positive if at least 40% overlaps between GFP and RFP objects). l

Input Red Cells!output Green Positive Red Cells.

l

Masking image Thresh GFP Well.

l

Remove depending on overlap 0.4 fraction (see Note 25).

(o) Measure Objects Intensity (see Note 26). l

Measure intensity of Red Cells in Aligned Green and Aligned Red images.

(p) Measure Objects Size Shape (see Note 27). l

Green Positive Red Cells.

l

Red Cells.

(q) Calculate Math (calculate fraction of positive cells). l

Calculate 100  count Green Positive Red Cells/ count Red Cells.

(r) Rescale Intensity. l

Rescale Crop Aligned Green!Rescaled Green for visualization.

(s) Rescale Intensity. l

Rescale Crop Aligned Red!Rescaled Red for visualization.

Methods for Quantitative Assessment of Protein Degradation

261

(t) Gray To Color. l

Input Rescaled Green, Rescaled Red!Color Image for visualization.

(u) Overlay Outlines. l

Outlines of objects in Green Positive Red Cells and Red Cells displayed on Color Image for visualization!Orig Overlay.

(v) Save Images. l

Save original image.

(w) Save Images. l

Save overlay image.

(x) Export To Spreadsheet. l Export analysis statistics to csv file. 3. The final exported file can be analyzed by plotting Math_Percent Green Positive column.

4

Notes 1. Cells should be ~70% confluent at time of transfection. 2. Different ratios of DNA:transfection reagent may need to be explored for optimal efficiency. 3. 6 μL FuGene HD can be used instead. 4. Transfection efficiency can be checked using the fluorescent microscope after 24 h. 5. The concentration of Hygromycin B can be increased to 100 μg/mL, after first round of 50 μg/mL to increase selection pressure. 6. Depending on the transfection efficiency, most of the cells will die and fall off the plate. 7. Antibiotic selection normally takes one or two week(s). In some cases, cells resistant to hygromycin will not be GFP or mCherry positive and further selection of GFP/mCherry double positive cells may be needed by FACS. 8. At this point, the transfection complexes and supernatant of the cells should be considered as infectious lentivirus. Use extra precautions to minimize infection hazards: (a) Double gloves and lab coat. (b) Have a container with bleach (~1 in. deep) in the hood. Anything touching transfection complexes/supernatant should be put into the bleach. When done in the hood, aspirate the bleach and add waste to normal waste streams.

262

Radosław P. Nowak et al.

9. Alternatively, the supernatant can be filtered with a 0.44 μm filter. 10. It is useful to transduce the cells at several concentrations of lentivirus (% of the final volume in the well) to prevent overinfection/multiple viral integrations per cell. Typical recommended concentrations are 10% and 50%, cells should be infected at 30–40% rate. 11. For HEK293T cells, use 5 μg/mL puromycin. 12. As alternative 30% of cells from confluent 10 cm dish can be used per 96-well plate. 13. For automated dispensing d300 (HP) dispenser is used at the Fischer laboratory with 10 mM stock solutions for compounds normalized to 0.5% DMSO across the plate. 14. To avoid aspirating the cells use p200 pipette tip appended to aspirating pipette. 15. The media FBS concentration can be reduced to 2% FBS or no FBS, to avoid possibility of clogging the nozzle of the flow analyzer. 16. Avoid making foam and air bubbles. Best to use 8-channel p200 pipette. 17. Easiest to load a previous worklist and adjust which wells will be read. Start by counting 2000 cells (i.e., 2000 “events” in worklist), adjust if more data points are necessary. 18. To increase dynamic range of the assay, it is important to increase the gain such that the slope of points is shifted to the upper right (increasing your signal). 19. User will need to press “Resume” for the run to start. Run for a full 96-well plate takes 60–90 min to complete. 20. No need to be too selective with the gate. 21. The GFP/mCherry ratio is multiplied by 10 for ease of analysis, but it can be any other factor, or 1. 22. Look through the data and note any outliers, for example low number of cells, or false-positive high/low mCherry values. 23. For time course imaging, most consistent results are achieved across the same 384-well plate. Include multiple incubation timepoints by treating cells in respective time interval with longest treatment treated first, and shortest last, following by readout. 24. The green channel can be dim and/or have big bright autofluorescent artifacts so this step helps to pull only the areas that look like cells. 25. This parameter changes sensitivity of the assay, higher the overlap the more stringent the assay, lower the overlap the more objects are accepted.

Methods for Quantitative Assessment of Protein Degradation

263

26. This part is not necessary and can be removed from analysis. Provides extra parameters to characterize cells by intensity of red and green channels. 27. This part is not necessary and can be removed from analysis. Provides extra parameters to characterize cells by their size.

Acknowledgments The authors gratefully acknowledge the generous financial support of the following sources: NIH grant NCI R01CA214608 (grant to E. S. F). We would like to thank members of Fischer and Ebert Lab at Dana-Farber Cancer Institute for useful discussions on development and optimization of these protocols. Conflict of Interest: E.S.F. is a founder, scientific advisory board (SAB), and equity holder in Civetta Therapeutics, Jengu Therapeutics (board member), and Neomorph, Inc. E.S.F. is an equity holder in C4 Therapeutics and a consultant to Novartis, Sanofi, EcoR1 capital, Deerfield and Astellas. The Fischer lab receives or has received research funding from Novartis, Ajax and Astellas. References 1. Salami J, Crews CM (2017) Waste disposal-an attractive strategy for cancer therapy. Science 355(6330):1163–1167. https://doi.org/10. 1126/science.aam7340 2. Winter GE et al (2015) DRUG DEVELOPMENT. Phthalimide conjugation as a strategy for in vivo target protein degradation. Science 348(6241):1376–1381. https://doi.org/10. 1126/science.aab1433 3. Lai AC, Crews CM (2017) Induced protein degradation: an emerging drug discovery paradigm. Nat Rev Drug Discov 16(2):101–114. https://doi.org/10.1038/nrd.2016.211 4. Wilkinson KD (2005) The discovery of ubiquitin-dependent proteolysis. Proc Natl Acad Sci 102(43):15280–15282. https://doi. org/10.1073/pnas.0504842102 5. Fischer ES et al (2014) Structure of the DDB1CRBN E3 ubiquitin ligase in complex with thalidomide. Nature 512(7512):49–53. https://doi.org/10.1038/nature13527 6. Bondeson DP et al (2015) Catalytic in vivo protein knockdown by small-molecule PROTACs. Nat Chem Biol 11(8):611–617. https://doi.org/10.1038/nchembio.1858

7. Riching KM et al (2018el) Quantitative livecell kinetic degradation and mechanistic profiling of PROTAC mode of action. ACS Chem Biol 13(9):2758–2770. https://doi. org/10.1021/acschembio.8b00692 8. Nowak RP et al (2018) Plasticity in binding confers selectivity in ligand-induced protein degradation. Nat Chem Biol 14(7):706–714. https://doi.org/10.1038/s41589-018-0055y 9. Gadd MS et al (2017) Structural basis of PROTAC cooperative recognition for selective protein degradation. Nat Chem Biol 13 (5):514–521. https://doi.org/10.1038/ nchembio.2329 10. Douglass EF, Miller CJ, Sparer G, Shapiro H, Spiegel DA (2013) A comprehensive mathematical model for three-body binding equilibria. J Am Chem Soc 135(16):6092–6099. https://doi.org/10.1021/ja311795d 11. Carpenter AE et al (2006) CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol 7(10):R100. https://doi.org/10.1186/gb-2006-7-10r100

Chapter 14 A High-Throughput Method to Prioritize PROTAC Intracellular Target Engagement and Cell Permeability Using NanoBRET James D. Vasta, Cesear R. Corona, and Matthew B. Robers Abstract Target engagement and cell permeation are important parameters that may limit the efficacy of proteolysistargeting chimeras (PROTACs). Here, we present an approach that facilitates both the quantitation of PROTAC binding affinity for an E3 ligase of interest, as well as the assessment of relative intracellular availability. We present a panel of E3 ligase target engagement assays based upon the NanoBRET Target Engagement platform. Querying E3 ligase engagement under live-cell and permeabilized-cell conditions allow calculation of an availability index that can be used to rank order the intracellular availability of PROTACs. Here we present examples where the cellular availability of PROTACs and their monovalent precursors are prioritized using NanoBRET assays for CRBN or VHL E3 ligases. Key words Target engagement, Intracellular affinity, BRET, NanoBRET™, Permeability, Intracellular Availability, E3 Ligase, PROTAC, Heterobifunctional degrader, Targeted protein degradation

1

Introduction Proteolysis-targeting chimeras (PROTACs) are bifunctional molecules that engage ubiquitin E3 ligases and induce degradation of intracellular proteins through a tightly regulated proteasomal mechanism [1].Although several successful PROTACs have been developed against a variety of intracellular target classes, the modular design of PROTACs as functional degraders faces a number of challenges. To achieve capacity as functional degraders, PROTACs encompass both E3 ligase- and target-binding moieties, generally bridged by aliphatic or polyethylene glycol-based linker moieties. Each of these functionalities can influence PROTAC efficacy in different ways. For instance, relatively minor structural variations can impact cell permeability and availability, formation of the ternary complex, or conformations that promote recognition by requisite degradation factors [2–5].

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_14, © Springer Science+Business Media, LLC, part of Springer Nature 2021

265

266

James D. Vasta et al.

Compared to its monovalent precursors, the resulting PROTAC is generally a high molecular weight inhibitor. As such, PROTACs often suffer from poor cell permeability and/or intracellular availability [2, 3], as well as limited capacity to engage the binding interfaces of both the E3 ligase and the target to be degraded. Consequently, the identification and optimization of cell-permeant PROTACs can be an iterative and challenging process. To date, methods of querying PROTAC permeability are limited to artificialized systems including PAMPA and Caco-2platforms, which are routinely used to characterize traditional small molecules. However, the sensitivity of these approaches is limited for PROTACs and other bifunctional inhibitors due to the low permeability and high molecular weight of such compounds [6, 7]. Newer methods that utilize reporter-tagged PROTAC derivatives (e.g., chloroalkanes) are attractive as they may offer more precise assessments of intracellular availability. However, such chemical reporter groups require an expanded synthetic work flow and may alter the properties of the molecule of interest. Therefore, the permeability properties of the modified entity may not reflect those of the parent compound [6]. Methods are desired to prioritize cellular availability of unmodified PROTACs as a key mechanistic readout. To aide in the process of PROTAC optimization, such methods should be scalable and compatible with intact, living cells. To enable a high-throughput readout for PROTAC cellular availability, we have developed an intracellular biosensor approach that utilizes E3 ligase engagement as a direct readout for PROTAC cellular entry. Toward this end we have developed a panel of NanoBRET™ target engagement (TE) assays for key E3 ligases or adapter proteins including CRBN and VHL (Fig. 1a). These methods enable a quantitative measure of target occupancy via competitive displacement of an energy transfer (NanoBRET) complex expressed in living cells [8–10]. Each complex is comprised of a NanoLuc-tagged target E3 ligase in dynamic equilibrium with a fluorescent reporter (NanoBRET™ Tracer) occupying a known ligand interface. Upon addition of the unmodified test compound, target engagement results in competitive displacement of the tracer and dose-dependent loss of the BRET signal. In a simple workflow, an availability index of PROTACs can be assessed by evaluating target engagement to the E3 ligase in both live and permeabilized cells (Fig. 1b). The resulting potency shift between intact and lysed cells, expressed as a ratio, reflects the intracellular availability of the PROTAC. Subsequent normalization of this potency shift to that of a control compound allows calculation of the availability index, which can be used to rank order PROTAC intracellular availability. This workflow can be executed in most HTS laboratories using a

A High-Throughput Method to Prioritize PROTAC Permeability

267

Fig. 1 A BRET-based approach to target engagement and assessment of relative intracellular availability. (a) Schematic representation of a BRET-based biosensor for E3 ligase engagement using BRET with NanoLuc (NLuc). Each complex is comprised of an NLuc-tagged target E3 ligase in dynamic equilibrium with a fluorescent reporter (tracer) occupying a known binding interface. Upon addition of the PROTAC or other ligand, target engagement results in competitive displacement of the tracer and dose-dependent loss of the BRET signal. (b) Schematic depiction of a workflow to evaluate the relative intracellular availability of PROTACs or other ligands. Binding potency is evaluated for ligands of interest in either live or permeabilized cells using the assay principle described in panel a, and the potency results for both conditions are mathematically combined to calculate an availability index that can be used to rank order the relative availability of the ligands

BRET-compatible luminometer. Here we present this method and analysis tool to determine availability indexes, with examples provided for CRBN- and VHL-based monovalent ligands and PROTACs.

268

2

James D. Vasta et al.

Materials

2.1 Transient Transfection of HEK293 Cells with NanoLuc® Fusions

1. HEK293 cells. 2. Miscellaneous tissue culture reagents and plasticware. 3. Cell culture medium: 90% (v/v) DMEM (Life technologies Cat# 11995), 10% (v/v) Fetal bovine serum. 4. Assay medium: Opti-MEM without phenol red (Life Technologies Cat#11058). 5. Transfection reagent (FuGENE® HD, Promega). 6. NanoLuc®-CRBN fusion vector (Promega). 7. DDB1 expression vector (Promega). 8. VHL-NanoLuc® fusion vector (Promega). 9. Promoterless DNA (Transfection Carrier DNA, Promega).

2.2 Determination of Compound Affinity in Live Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)

1. HEK293 cells transfected with desired E3 ligase according to the Transient Transfection of HEK293 Cells with NanoLuc® Fusions section above. 2. Cell culture medium: 90% (v/v) DMEM (Life technologies Cat# 11995), 10% (v/v) fetal bovine serum. 3. White, nonbinding surface (NBS) 96-well plates, appropriately coated to avoid adsorptive properties of the fluorescent tracers (Corning Cat# 3600). 4. NanoBRET™ Tracer Dilution Buffer (Promega). 5. Polypropylene plastic ware (buffer troughs, serial dilution troughs, etc.). 6. Assay medium: Opti-MEM without phenol red (Life Technologies Cat#11058). 7. Luciferase substrate (NanoBRET™ Nano-Glo® Substrate, Promega). 8. Dimethylsulfoxide (DMSO). 9. CRBN Tracerstock solution: 400μM in DMSO (stored at
80  C; stable to at least 5 freeze–thaw cycles). 10. VHL Tracerstock solution: 400μM in DMSO (stored at
80  C; stable to at least 5 freeze–thaw cycles). 11. Test compound stock solution: 1000 desired final concentration in DMSO. 12. Test compound working solution (10): 1 part (by volume) test compound stock solution diluted with 99 parts (by volume) assay medium (see Note 1).

A High-Throughput Method to Prioritize PROTAC Permeability

269

Fig. 2 Verification data sets for the CRBN E3 ligase target engagement assay in either live or permeabilized cells. (a) Determination of apparent tracer affinity (left) and the impact of tracer concentration (right) for CRBN E3 ligase in live HEK293 cells. Coexpression of the DDB1 protein improves both assay window and assay stability (data not shown). The tracer concentration chosen for analysis (0.5 μM) is highlighted in red (right). (b) The same validation approach was applied to optimize a CRBN assay in digitonin-permeabilized HEK293 cells, using an equilibration time of 10 min

13. BRET-compatible luminometer equipped with 450 nm (bandpass) and 600 nm (longpass) filters. (e.g., Glomax Discover, PerkinElmer EnVision, or BMG Clariostar). 14. Extracellular NanoLuc® Inhibitor (Promega). 15. 100 live-mode CRBN Tracer reagent: 50μM CRBN Tracer in DMSO prepared from the CRBN Tracer stock solution (see Note 2 and Fig. 2). 16. 20 live-mode CRBN Tracer reagent: 1 part 100 live-mode CRBN Tracer reagent, 4 parts NanoBRET Tracer Dilution Buffer (see Note 3). 17. 100 live-mode VHL Tracer Reagent: 100μM VHL Tracer in DMSO prepared from the VHL Tracer stock solution (see Note 2 and Fig. 3).

270

James D. Vasta et al.

Fig. 3 Verification data sets for the VHL E3 ligase target engagement assay in either live or permeabilized cells. (a) Determination of apparent tracer affinity (left) and the impact of tracer concentration (right) for VHL E3 ligase in live HEK293 cells. The tracer concentration chosen for analysis (1 μM) is highlighted in red (right). (b) The same validation approach was applied to optimize a VHL assay in digitonin-permeabilized HEK293 cells, using an equilibration time of 25 min

18. 20 live-mode VHL Tracer Reagent: 1 part 100 live-mode VHL Tracer reagent, 4 parts NanoBRET Tracer Dilution Buffer (see Note 3). 19. 3 complete live-mode detection solution: 1:166 dilution of the NanoBRET™ Nano-Glo® substrate plus a 1:500 dilution of the Extracellular NanoLuc® Inhibitor in assay medium (see Note 4). 2.3 Determination of Compound Affinity in Permeabilized Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)

1. Reagent and equipment 1–13 described in Subheading 2.2. 2. Digitonin stock solution: 50 mg/mL digitonin in DMSO. 3. 10 permeabilization reagent: 1 part (by volume) digitonin stock solution plus 99 parts assay medium. 4. 100 permeabilized-mode CRBN Tracer reagent: 13μM CRBN Tracer in DMSO prepared from the CRBN Tracer stock solution (see Note 2 and Fig. 2).

A High-Throughput Method to Prioritize PROTAC Permeability

271

5. 20 permeabilized-mode CRBN Tracer reagent: 1 part 100permeabilized-mode CRBN Tracer reagent, 4 parts NanoBRET Tracer Dilution Buffer (see Note 3). 6. 100 permeabilized-mode VHL Tracer reagent: 25μM VHL Tracer in DMSO prepared from the VHL Tracer stock solution (see Note 2 and Fig. 3). 7. 20 permeabilized-mode VHL Tracer reagent: 1 part 100 permeabilized-mode VHL Tracer reagent, 4 parts NanoBRET Tracer Dilution Buffer (see Note 3). 8. 3 complete permeabilized-mode detection solution: 1:166 dilution of the NanoBRET™ Nano-Glo® substrate in assay medium (see Note 4).

3

Methods

3.1 Transient Transfection of HEK293 Cells with NanoLuc® Fusions

Transient transfection is a convenient and robust way to introduce E3 ligase/NanoLuc® fusions into various cell types, such as the HEK293 cells used in this target engagement-based approach. In this section, we provide example protocols for the transient transfection of HEK293 cells with VHL or CRBN E3 ligases. These protocols should be generalizable to other E3 ligases when working with HEK293 cells, and could be modified to support application to other cell types if desired. Alternatively, if the user plans to utilize a cell line that is stably expressing the target/NanoLuc® fusion of interest, skip this section and proceed to Subheadings 3.2 and/or 3.3. 1. Cultivate HEK293 cells appropriately prior to assay and resuspend cells into a single-cell suspension using cell culture medium. 2. Adjust the cell density to 2  105 cells/mL in cell culture medium in a sterile conical tube. 3. Prepare desired E3 ligase plasmid DNAs for transfection in sterile conical tubes: (a) Example: for VHL E3 Ligase, combine by mass 1 part VHL-NanoLuc® fusion vector and 9 parts promoterless DNA in assay medium at a total DNA concentration of 10μg/mL. (b) Example: for CRBN E3 Ligase, combine by mass 1 part NanoLuc®-CRBN fusion vector and 9 parts DDB1 expression vector in assay medium at a total DNA concentration of 10μg/mL (see Note 5). 4. Initiate transfection complex formation by adding 3 parts (by volume) FuGENE® HD for each part (by mass) of DNA.

272

James D. Vasta et al.

Mix by inversion five times. Allow 10–15 min for complexation to occur (see Note 6). 5. Combine one part (by volume) of transfection complex with 20 parts (by volume) HEK293 cell suspension at 2  105 cells/ mL in a sterile conical tube. Mix gently by inversion five times. 6. Dispense cells/lipid-DNA complex into a sterile tissue culture flask and incubate at least 20 h to allow expression to occur. We recommend a cell density of approximately 55,000–80,000 cells/cm2 during the transfection (for example, use approximately 4–6 million cells for a T75 flask). Larger or smaller bulk transfections should be scaled accordingly using this ratio. 3.2 Determination of Compound Affinity in Live Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)

This section provides a protocol for the live-cell mode of the NanoBRET™ target engagement assay. It can be used as a standalone assay to evaluate the binding potency of test compounds or PROTACs in live cells to either VHL or CRBN E3 ligases, or in conjunction with Subheadings 3.3 and 3.4 to evaluate the relative intracellular availability of the test compound(s). Verification data supporting the recommended concentrations of NanoBRET™ Tracer are provided in Figs. 2 and 3. This protocol should also be adaptable to other E3 ligases with further assay development (e.g., development and characterization of an appropriate tracer for the E3 ligase of interest) [9]. 1. Harvest transfected cells via trypsinization and resuspend in cell culture medium. 2. Centrifuge the cells and resuspend the pellet in assay medium at 2  105 cells/mL. 3. Dispense 85μL cell suspension per well into white, 96-well NBS plates (see Note 7). 4. Add 5μL per well of the 20 live-mode VHL Tracer reagent, 20 live-mode CRBN Tracer reagent, or an alternative 20 live-mode tracer reagent depending on the E3 ligase being tested (see Notes 8 and 9). 5. Mix on an orbital shaker for 15 s at 900 rpm (see Note 10). 6. Add 10μL per well of the 10 test compound working solution. 7. Mix on an orbital shaker for 15 s at 900 rpm. 8. Incubate the plate at 37  C and 5% CO2 for 1 h 45 min. 9. Allow plate to re-equilibrate to ambient temperature for approximately 15 min. 10. Add 50μL of 3 complete live-mode detection solution and incubate at ambient temperature for 3 min to allow production of luminescence.

A High-Throughput Method to Prioritize PROTAC Permeability

273

11. Measure donor emission (450 nm) and acceptor emission (600 nm) using the BRET compatible luminometer. 12. Generate raw BRET ratio values by dividing the acceptor emission value by the donor emission value. 13. Convert raw BRET ratio values to milliBRET units (mBU) by multiplying each raw BRET ratio value by 1000. 14. (Optional) If a background correction is desired, use the NanoBRET Equation (Eq. 1).   Acceptor sample Acceptor no tracer control  1, 000 ð1Þ
Donor sample Donor no tracer control 15. Determine the potency (IC50) of your test compound by plotting the BRET ratio values (as mBUs) as a function of test compound concentration and fitting to the sigmoidal doseresponse (variable slope) equation (Eq. 2).   Y ¼ Bottom þ ðTop
Bottom Þ= 1 þ 10ððLog ðIC 50 Þ
X Þ∗HillSlope Þ ð2Þ where X ¼ test compound concentration and Y ¼ BRET ratio. 3.3 Determination of Compound Affinity in Permeabilized Cells for E3 Ligases (CRBN and VHL E3 Ligase Examples)

This section provides a protocol for the permeabilized mode of the NanoBRET™ target engagement assay. It can be used as a standalone assay to evaluate the binding potency of test compounds or PROTACs under cell-free conditions to either VHL or CRBN E3 ligases, or in conjunction with Subheadings 3.2 and 3.4 to evaluate the relative intracellular availability of the test compound(s). Verification data supporting the recommended concentrations of NanoBRET™ Tracer are provided in Figs. 2 and 3. This protocol should also be adaptable to other E3 ligases with further assay development (e.g., development and characterization of an appropriate tracer for the E3 ligase of interest) [9]. 1. Harvest transfected cells via trypsinization and resuspend in cell culture medium. 2. Centrifuge the cells and resuspend the pellet in assay medium at 2  105 cells/mL. 3. Dispense 75μL cell suspension per well into white, 96-well NBS plates (see Note 7). 4. Add 5μL per well of the 20 permeabilized-mode VHL Tracer reagent, 20 permeabilized-mode CRBN Tracer reagent, or alternative 20 permeabilized-mode tracer reagent depending on the E3 ligase being tested (see Notes 8 and 9). 5. Mix on an orbital shaker for 15 s at 900 rpm (see Note 10).

274

James D. Vasta et al.

6. Add 10μL per well of the 10 test compound working solution. 7. Mix on an orbital shaker for 15 s at 900 rpm. 8. Add 10μL 10 permeabilization reagent. 9. Mix on an orbital shaker for 15 s at 900 rpm. 10. Incubate plate at ambient temperature protected from light for a consistent period of time for all assay plates, with a maximum incubation of 25 min (see Note 11). 11. Add 50μL of 3 complete permeabilized-mode detection solution and incubate at ambient temperature for 0.5–1 min to allow production of luminescence. 12. Measure donor emission (450 nm) and acceptor emission (600 nm) using the BRET compatible luminometer. 13. Generate raw BRET ratio values by dividing the acceptor emission value by the donor emission value. 14. Convert raw BRET ratio values to milliBRET units (mBU) by multiplying each raw BRET ratio value by 1000. 15. (Optional) If a background correction is desired, use the NanoBRET Equation (Eq. 1). 16. Determine the potency (IC50) of your test compound by plotting the BRET ratio values (as mBUs) as a function of test compound concentration and fitting to the sigmoidal doseresponse (variable slope) equation (Eq. 2). 3.4 Assessment of Relative Intracellular Availability for E3 Ligase Ligands Via the Availability Index

The plasma membrane “permeability” of a molecule encompasses many parameters, including the ability to traverse the membrane by passive diffusion, the rate at which that passive diffusion process occurs, the rates of active import and efflux processes, and the concentration gradient of the molecule that is established on either side of the membrane after equilibrium is achieved. The last parameter, also known as “intracellular availability”, is a recently reported drug disposition parameter that is typically challenging to measure, but would be enabling in lead optimization workflows [11]. Herein, we describe a simple approach that uses the change in apparent potency of a test compound in live cells versus permeabilized cells to evaluate the relative availability of that molecule across the plasma membrane, and apply that approach to PROTACS and other E3 ligase ligands. The assay design is based on the premise that ligands with decreasing intracellular availability (i.e., a lower concentration inside cells compared to that present in the medium) will show a commensurate shift toward weaker target engagement potency as determined in live cells (Kd,live) compared its intrinsic affinity as determined under cell-free conditions (Kd). Conversely, a highly available compound that establishes an identical free concentration

A High-Throughput Method to Prioritize PROTAC Permeability

275

on either side of the plasma membrane at equilibrium should demonstrate equivalent on-target potency. Evaluation of ligand availability using this approach relies on a few key assumptions (similar to those established previously [11]) regarding the system under study (i.e., a cell population containing a ligand of interest): 1. At equilibrium, there will be a fixed ratio (Fpt) of the free concentration of ligand inside the cell (proximal to the target) and outside the cell according to Eq. (3). ½L pt ¼ F pt ½L medium

ð3Þ

where [L]pt equals the free concentration of the ligand inside the cell and proximal to the target and [L]medium equals the free concentration of the ligand in the medium. 2. The total amount of free ligand added to the system (Ltotal) is much greater than the amount of ligand absorbed by the cells. Hence: ½L medium ffi ½L total

ð4Þ

3. The behavior of the target of interest under “cell-free” conditions (i.e., in a population of permeabilized cells or a lysate) is identical to its behavior inside the cell. 4. The behavior of the ligand of interest under “cell-free” conditions is identical to its behavior inside the cell. 5. The fractional occupancy (R) of the target by the ligand of interest is reflected by the Langmuir isotherm [12] (Eq. (5)). R¼

½L total K d þ ½L total

ð5Þ

Substituting Eqs. (3) and (4) into Eq. (5), we find that the cellular fractional occupancy for the target (Rcell) is reflected by Eq. (6). Rcell ¼

½L pt F pt  ½L total ½L total ¼ ¼ K d þ ½L pt K d þ F pt  ½L total FK d þ ½L total

ð6Þ

pt

In this way, the apparent live-cell potency of the ligand according to Eq. (6) is simply adjusted for the intracellular availability of the compound proximal to the target, as defined in Eq. (7). K d,cell ¼

Kd F pt

ð7Þ

Rearranging Eq. (7), we define the relative binding affinity (RBA) of the ligand according to Eq. (8), providing a relationship by which intracellular availability proximal to the target can be

276

James D. Vasta et al.

related to the change in binding potency measured in live cells relative to the intrinsic affinity of the ligand for the target: RBA ¼

K d,cell 1 ¼ F pt Kd

ð8Þ

According to assumptions 3 and 4, a compound that establishes identical-free concentrations inside and outside the cell will demonstrate an RBA value of 1. We have observed RBA values close to 1 for monovalent VHL and CRBN ligands (vide infra), suggesting that assumption 3 above is approximately valid for these E3 ligases under our assay conditions. However, while we require assumption 4 to hold true, assumption 3 may not be completely satisfied for some targets where binding behavior may actually change in live versus permeabilized cell conditions (e.g., changes in target complexation status, changes in levels of competing intracellular metabolites, or changes due to the permeabilization conditions). To account for this problem, we expand this approach to include a permeable control compound (see Note 12) that serves as a benchmark for comparison and calibration of the RBA value. Thus, inclusion of a permeable control compound makes this approach applicable to many target classes, even when assumption 3 is not valid (e.g., for target classes like kinases where levels of intracellular ATP may cause right-shifted pharmacology in cells compared to cell-free conditions). In practice, the potency of a compound series is determined in both live and permeabilized conditions, where the potency under permeabilized cell conditions is a proxy for intrinsic affinity (Kd) of the interaction. The RBA value is estimated using Eq. (9): RBA ¼

K d,cell Potency live
mode ffi Kd Potency permeabilized
mode

ð9Þ

The RBA for the permeable control compound is used to calibrate the assay behavior, and the RBA values for all compounds are then normalized to that of the permeable control compound to establish an availability index (AI). For example, demonstration of measurable live-cell target engagement potency (e.g., a measurable AI value) supports the ability of the compound to traverse the membrane. Secondly, the actual availability index value reflects the relative target proximal concentration of the compound compared to the control, where AI values greater than 1 suggest reduced availability compared to the control, and vice versa. Examples of this approach as applied to the E3 ligases VHL and CRBN are demonstrated in Figs. 4 and 5, respectively. For VHL, monovalent ligand VH298 was used as the control (AI ¼ 1.0). VH298 also demonstrated an RBA value of 0.7, supporting the validity of assumption 3 above for VHL. After a 2-h incubation,

A High-Throughput Method to Prioritize PROTAC Permeability

277

Fig. 4 Example target engagement and availability profiling for VHL ligands. A panel of VHL ligands and PROTACs (a) were evaluated for potency in live (b) and permeabilized (c) HEK293 cells. Individual data points represent the mean of quadruplicate measurements. VHL ligand potency and availability parameters are tabulated in panel (d). VH298 was used as the permeable control compound for comparison. PROTAC ARV-771 shows better intracellular availability (AI ¼ 2.4) compared to that of PROTAC MZ1 (AI ¼ 8.7) at 2 h

PROTAC ARV-711 showed both improved binding potency and availability (AI ¼ 2.4) compared to PROTAC MZ1 (AI ¼ 8.7). For CRBN, the monovalent ligand lenalidomide was used as the permeable control (AI ¼ 1.0). Moreover, the RBA value for lenalidomide was 0.63, reasonably close to a value of 1. Iberdomide

278

James D. Vasta et al.

Fig. 5 Example target engagement and availability profiling for CRBN ligands. A panel of CRBN ligands and PROTACs (a) were evaluated for potency in live (b) and permeabilized (c) HEK293 cells. Individual data points represent the mean of quadruplicate measurements. CRBN ligand potency and availability parameters are tabulated in panel (d). Iberdomide demonstrated similar intracellular availability compared to the permeable control lenalidomide. PROTAC dBET6 showed improved availability (AI ¼ 5.6) compared to that of the linker variant dBET1 (AI ¼ 60)

demonstrated a similar but slightly improved availability (AI ¼ 0.79) compared to lenalidomide, confirming the high availability of this ligand. After a 2 h incubation, PROTAC dBET6 demonstrated significantly greater availability (AI ¼ 5.6) compared

A High-Throughput Method to Prioritize PROTAC Permeability

279

to linker variant dBET1 (AI ¼ 60), further emphasizing the importance of linker optimization in the development of efficacious bifunctional degraders. Though intracellular availability and permeability (rate of entry) are distinct parameters, we note that the rank order of AI scores for VHL ligands agreed well with the expected permeability rank order according to previously reported PAMPA [13, 14], MDCK-MDR1 [14], or Caco-2 values [6]. In the case of CRBN PROTACs dBET1 and dBET6, rank order in the AI scores was also in agreement with the permeability rank order expected from previous measurements using the Caco-2 assay [15]. However, we note that while intracellular availability and permeability rate may correlate in some models, these two parameters are independent and not expected to correlate a priori. Thus, additional approaches to directly measure the rate of cellular entry or “intracellular permeability” for low permeability molecules like PROTACs are desirable. 1. Determine the potency of your test compound to the E3 ligase of interest in live-mode according to Subheading 3.2 and in permeabilized-mode according to Subheading 3.3 (see Note 12). 2. Calculate the relative binding affinity (RBA) of the compound using Eq. (9) (see Note 13). 3. Calculate the availability index (AI) of the compound using Eq. (10) AI ¼

RBAtest compound RBApermeable control

ð10Þ

4. Compare the AI values for your test compounds to that of the permeable control compound (AI value of 1, see Note 14).

4

Notes 1. For preparation of test compound dilution series, we recommend first preparing a 1000X dilution series of the test compound in DMSO, after which 1 part (by volume) of this 1000X series can be diluted with 99 parts (by volume) assay medium to prepare a 10X test compound dilution series. This ensures that the same amount of DMSO is present in all assay wells. 2. The final concentration of NanoBRET™ tracers to use in the assay (and hence the concentration of the 100X stock solution) is determined by standard assay verification procedures described previously [9]. Briefly, the affinity of the tracer for the NanoLuc fusion and the impact of the tracer concentration on test compound pharmacology are evaluated under the

280

James D. Vasta et al.

conditions of study (e.g., in live or permeabilized cells). The working tracer concentration is then chosen such that an adequate assay window is provided without causing a significant right shift in the pharmacology for the test compound. This tracer concentration is often between EC50 and EC80 of the tracer. Example verification data for the CRBN and VHL assays are provided in Figs. 2 and 3, respectively. More accurate quantitation can be achieved by reducing the tracer concentration, provided that an adequate assay window is achieved. 3. Prepare 20 NanoBRET tracer reagents on the day of assay, ideally within 5 h of addition to assay plates. Discard 20 NanoBRET tracer reagents after use and do not store. 4. Prepare 3 Complete Detection Solutions immediately prior to BRET measurement and use within 1 h. Make sure to add the Extracellular NanoLuc® Inhibitor component when preparing the live-mode reagent and withhold the Extracellular NanoLuc® Inhibitor when preparing the permeabilized-mode reagent. 5. For CRBN, it has been found the coexpression of the DDB1 partner protein significantly improves assay performance and stability. Coexpression of a partner protein may be a useful assay design parameter to consider for other E3 ligases as needed for assay performance. 6. Add the FuGENE HD directly to the cell suspension and do not allow it to touch the sides of the conical tube. 7. Periodically mix cells appropriately to avoid settling of the cell suspension. This could be by inversion of a conical tube or by gentle pipetting to mix. 8. Due to the viscosity of the 20 NanoBRET tracer reagents, dispense these solutions slowly. 9. If studying an E3 ligase other than CRBN or VHL where the formulation of the 100 and 20 tracer reagents are provided herein, the appropriate final concentration of tracer to use in the assay should be determined through standard assay verification workflows. That includes determination of the apparent intracellular affinity of the tracer and the impact of the tracer on the pharmacology of an unlabeled test compound. Assay design verification has been discussed in depth previously [9]. 10. Mixing may vary between orbital shakers and should be optimized for each individual unit accordingly by visual inspection. A homogenous suspension should be achieved with complete dispersion of the viscous tracer reagent. 11. After permeabilization, the VHL and CRBN assays show performance drifts that include time-dependent right-shifting of test compound potency and time-dependent reduction of the

A High-Throughput Method to Prioritize PROTAC Permeability

281

assay window. This phenomenon is not fixed using alternative permeabilization reagents, changing expression conditions, or adding stabilizer reagents, and is likely due to dissociation of the E3 ligase complexes over time. For best performance of the permeabilized mode of the assay, we recommend a consistent incubation time for all assay plates after permeabilization and prior to BRET measurements (5 min), with a maximum incubation time of 25 min. Moreover, when assessment of relative intracellular availability is desired, we recommend including a permeable control compound in both the live and permeabilized mode measurements so that it can be used to establish an availability index for comparison (see Subheading 3.4). 12. When evaluating relative availability, it is important to include a permeable control compound to establish an RBA value for comparison. The permeability of this compound may have been assessed by conventional approaches including PAMPA or Caco-2 assays, but it should otherwise be able to access the target inside the cell. Once established, the RBA of the permeable control compound can then be used for comparison to other compounds to evaluate the availability index (AI) as the RBA value for the test compound normalized to that of the permeable control. See Subheading 3.4 for more details. 13. The RBA value should theoretically reflect the intracellular availability of the compound proximal to the target. Ideally, the RBA for a permeable compound that is not otherwise modified by the cellular milieu and establishes an equivalent concentration on either side of the membrane will be a value of 1, indicating that the potency of the compound for the E3 ligase is the same regardless of the presence or absence of the membrane. However, due to assay behavioral differences between the live-mode and permeabilized-mode assays, the RBA for a permeable compound may be different from a value of 1. Thus, a permeable control compound should be used to establish the RBA that will serve as a benchmark for comparison. 14. Compounds with similar intracellular availability to the control compound will show an AI of 1. The AI for compounds with reduced availability compared to the control compound will increase, with the AI being incalculable for completely impermeable compounds where live-mode binding is not detected. AI values less than 1 would indicate that the compound accumulates in cells to higher relative concentrations compared to that achieved by the chosen control compound.

282

James D. Vasta et al.

References 1. Shapira M, Calabrese MF, Bullock AN, Crews CM (2019) Targeted protein degradation: expanding the toolbox. Nat Rev Drug Discov 18(12):949–963. https://doi.org/10.1038/ s41573-019-0047-y 2. Daniels DL, Riching KM, Urh M (2019) Monitoring and deciphering protein degradation pathways inside cells. Drug Discov Today Technol 31:61–68. https://doi.org/10. 1016/j.ddtec.2018.12.001 3. Liu J, Ma J, Xia J, Li Y, Wang ZP, Wei W (2020) PROTACs: a novel strategy for cancer therapy. Semin Cancer Biol 67(Pt 2):171–179. https://doi.org/10.1016/j.semcancer.2020. 02.006 4. Maple HJ, Clayden N, Baron A, Stacey C, Felix R (2019) Developing degraders: principles and perspectives on design and chemical space. MedChemComm 10(10):1755–1764. https://doi.org/10.1039/c9md00272c 5. Soares P, Gadd MS, Frost J, Galdeano C, Ellis L, Epemolu O, Rocha S, Read KD, Ciulli A (2018) Group-Based Optimization of Potent and Cell-Active Inhibitors of the Von Hippel-Lindau (VHL) E3 Ubiquitin Ligase: Structure-Activity Relationships Leading to the Chemical Probe (2S,4R)-1-((S)-2-(1-Cyanocyclopropanecarboxamido)-3,3-dimethylbutanoyl)-4-hydroxy-N-(4-(4-methylthiazol5-yl)benzyl)pyrrolidine-2-carboxamide (VH298). J Med Chem 61(2):599–618. https://doi.org/10.1021/acs.jmedchem. 7b00675 6. Foley CA, Potjewyd F, Lamb KN, James LI, Frye SV (2020) Assessing the cell permeability of bivalent chemical degraders using the chloroalkane penetration assay. ACS Chem Biol 15 (1):290–295 7. Zoppi V, Hughes SJ, Maniaci C, Testa A, Gmaschitz T, Wieshofer C, Koegl M, Riching KM, Daniels DL, Spallarossa A, Ciulli A (2019) Iterative design and optimization of initially inactive proteolysis targeting chimeras (PROTACs) identify VZ185 as a potent, fast, and selective von hipper-Lindau (VHL) based dual degrader probe of BRD9 and BRD7. J Med Chem 62(2):699–726 8. Robers MB, Dart ML, Woodroofe CC, Zimprich CA, Kirkland TA, Machleidt T, Kupcho KR, Levin S, Hartnett JR, Zimmerman K, Niles AL, Ohana RF, Daniels DL, Slater M, Wood MG, Cong M, Cheng YQ, Wood KV (2015) Target engagement and drug residence time can be observed in living cells with BRET. Nat Commun 6:10091. https://doi.org/10. 1038/ncomms10091

9. Robers MB, Vasta JD, Corona CR, Ohana RF, Hurst R, Jhala MA, Comess KM, Wood KV (2019) Quantitative, real-time measurements of intracellular target engagement using energy transfer. Methods Mol Biol 1888:45–71. https://doi.org/10.1007/978-1-4939-88914_3 10. Vasta JD, Corona CR, Wilkinson J, Zimprich CA, Hartnett JR, Ingold MR, Zimmerman K, Machleidt T, Kirkland TA, Huwiler KG, Ohana RF, Slater M, Otto P, Cong M, Wells CI, Berger BT, Hanke T, Glas C, Ding K, Drewry DH, Huber KVM, Willson TM, Knapp S, Muller S, Meisenheimer PL, Fan F, Wood KV, Robers MB (2018) Quantitative, wide-Spectrum kinase profiling in live cells for assessing the effect of cellular ATP on target engagement. Cell Chem Biol 25(2):206–214. e211. https:// doi.org/10.1016/j.chembiol.2017.10.010 11. Mateus A, Treyer A, Wegler C, Karlgren M, Matsson P, Artursson P (2017) Intracellular drug bioavailability: a new predictor of system-dependent drug disposition. Sci Rep 7:43047. https://doi.org/10.1038/ srep43047 12. Hulme EC, Trevethick MA (2010) Ligand binding assays at equilbrium: validation and interpretation. Br J Pharmacol 161:1219–1237 13. Frost J, Galdeano C, Soares P, Gadd MS, Grzes KM, Ellis L, Epemolu O, Shimamura S, Bantscheff M, Grandi P, Read KD, Cantrell DA, Rocha S, Ciulli A (2016) Potent and selective chemical probe of hypoxic signaling downstream of HIF-α hydroxylation via VHL inhibition. Nat Commun 7:13312. https:// doi.org/10.1038/ncomms13312 14. Raina K, Lu J, Qian Y, Altieri M, Gordon D, Rossi AMK, Wang J, Chen X, Dong H, Siu K, Winkler JD, Crew AP, Crews CM, Coleman KG (2016) PROTAC-induced BET protein degradation as a therapy for castration-resistant prostate cancer. Proc Natl Acad Sci U S A 113 (26):7124–7129. https://doi.org/10.1073/ pnas.1521738113 15. Winter GE, Mayer A, Buckley DL, Erb MA, Roderick JE, Vittori S, Reyes JM, Iulio JD, Souza A, Ott CJ, Roberts JM, Zeid R, Scott TG, Paulk J, Lachance K, Olson CM, Dastjerdi S, Bauer S, Lin CY, Gray NS, Kelliher MA, Churchman LS, Bradner JE (2017) BET bromodomain proteins function as master transcription elongation factors independent of CDK9 recruitment. Mol Cell 67(1):5–18. https://doi.org/10.1016/j.molcel.2017.06. 004

Chapter 15 Profiling CELMoD-Mediated Degradation of Cereblon Neosubstrates Joel W. Thompson, Thomas Clayton, Gody Khambatta, Leslie A. Bateman, Christopher W. Carroll, Philip P. Chamberlain, and Mary E. Matyskiela Abstract Targeted protein degradation is garnering increased attention as a therapeutic modality due in part to its promise of modulating targets previously considered undruggable. Cereblon E3 Ligase Modulating Drugs (CELMoDs) are one of the most well-characterized therapeutics employing this modality. CELMoDs hijack Cereblon E3 ligase activity causing neosubstrates to be ubiquitinated and degraded in the proteasome. Here, we describe a suite of assays—cellular substrate degradation, confirmation of CELMoD mechanism of action, in vitro ubiquitination, and Cereblon binding—that can be used to characterize CELMoD-mediated degradation of Cereblon neosubstrates. While the assays presented herein can be run independently, when combined they provide a strong platform to support the discovery and optimization of CELMoDs and fuel validation of targets degraded by this drug modality. Key words Cereblon, CELMoD, Targeted protein degradation, In vitro ubiquitination, Cellular neosubstrate degradation, Cereblon binding

1 1.1

Introduction Overview

Targeted protein degradation has emerged as an exciting new modality in drug discovery. In this approach, a protein of interest is recruited to an ubiquitin ligase (E3) enzyme using a mole, leading to its ubiquitination and subsequent proteasomal degradation [1]. Two types of compounds have thus far been utilized for targeted protein degradation. Heterobifunctional ligands employ two independent binding moieties—one that binds to an E3 enzyme and a second that binds to the target of interest— connected through a linker to recruit targets to an E3 enzyme for ubiquitination [2, 3]. Alternatively, ligands that utilize a “molecular glue” mechanism do not typically bind independently to the target. Instead, a protein interaction surface formed on the ligand: E3 ligase complex provides a docking site for target protein binding [4]. The best understood ligands that function as molecular glues

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_15, © Springer Science+Business Media, LLC, part of Springer Nature 2021

283

284

Joel W. Thompson et al.

are the FDA approved Cereblon E3 Ligase Modulating Drugs (CELMoDs) thalidomide and its derivatives lenalidomide and pomalidomide, which bind to the Cereblon substrate receptor subunit of a CUL4-DDB1-Rbx1 E3 ligase complex (CRL4Cereblon) [5–7]. Mechanism of action studies identified the Cereblon neosubstrates (proteins regulated by Cereblon only in the presence of CELMoD) Aiolos and Ikaros as critical efficacy targets of CELMoDs in hematological cancers [8–10]. However, the synthesis of expanded CELMoD libraries based on thalidomide, combined with structural studies of Cereblon/CELMoD/neosubstrate complexes, mass spectrometry, and reporter assays, has revealed that a much larger set of proteins with diverse roles in normal and disease biology can be targeted for degradation by CELMoDs [11–16]. As research into the potential of CELMoDs in drug discovery continues, it is likely that the list of proteins that can be targeted for degradation by this modality will continue to grow. Here, we detail an experimental platform designed to evaluate new CELMoDs and validate potential Cereblon neosubstrates of these ligands. While these assays are described in detail for interrogating CELMoDs, it is important to note that these same assays can be used to investigate Cereblon-based heterobifunctional ligands. Moreover, while these assays can be run individually, when combined they establish a solid foundation to support the discovery and optimization of active CELMoDs and fuel the validation of targets degraded by this class of compounds. 1.2 High-Throughput Substrate Degradation

CELMoD-mediated Cereblon neosubstrate degradation can be monitored by measuring levels of the neosubstrate protein in cells. One simple and high-throughput way to do this is via an enzyme fragment complementation (EFC) assay. In these assays, a small fragment of a reporter enzyme is appended onto the target protein of interest. A cell line expressing the target-reporter enzyme fragment fusion protein can be treated with CELMoDs. Following incubation with CELMoD cells are lysed with an assay mixture including the larger fragment of the reporter enzyme. When the large fragment and the small fragment of the reporter interact, this creates an active enzyme leading to a luminescent signal proportional to the level of target protein present in the lysate. If CELMoD treatment promotes target degradation, this can be detected via a decreased signal in the EFC assay. HiBiT and enhanced ProLabel (ePL) are two commercially available systems for performing EFC assays [17, 18]. Here we detail how to generate a stable cell line expressing an ePL-tagged protein of interest, how to optimize the EFC assay, and how to perform a dose response curve assay to determine Ymin and DC50 values for a tested CELMoD against a Cereblon neosubstrate. DC50 is the concentration of drug that gives half-maximal degradation of the substrate and Ymin is the minimum percentage of substrate remaining following drug

CELMoD Directed Proteolysis of CRBN Neosubstrates

285

treatment. Comparing depth of degradation and potency are important parameters for optimization of CELMoDs against a specific neosubstrate. Thus, the substrate degradation assay presented below is straightforward, quantitative, and can also be scalable for screening compound libraries. 1.3 Chemical and Genetic Approaches to Confirm CELMoD MOA

Next, a set of experiments using a combination of small molecule and genetic tools to confirm CELMoD cellular mechanism of action (MOA) against a putative Cereblon neosubstrate identified from an EFC assay is described. Specifically, how to generate a Cereblon knockout (KO) cell line via CRISPR/Cas9 and how to use this cell line to confirm the requirement of Cereblon expression for cellular degradation of an endogenous target of interest via immunoblotting are detailed. Subsequently, an approach to co-treating cells with CELMoD and either the proteasome inhibitor, MG132, or the NEDD8 E1 inhibitor, MLN4924, to confirm cellular MOA via immunoblotting is presented. If a CELMoD is unable to reduce neosubstrate protein levels in the presence of MG132, this would suggest the CELMoD is working via a proteasome-dependent mechanism to regulate the neosubstrate. Similarly, if cells are treated with MLN4924, which blocks activation of all cullin-RING ubiquitin ligase complexes, a CELMoD should be incapable of facilitating degradation of the target of interest confirming the requirement of CRL4Cereblon for CELMoD activity.

1.4 In Vitro Ubiquitination

A biochemical system is presented describing an in vitro ubiquitination assay to determine whether neosubstrates are ubiquitinated by CRL4Cereblon in a CELMoD-dependent manner. To assay the direct CELMoD-dependent covalent attachment of ubiquitin to a target, CELMoD and purified CRL4Cereblon, ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzyme (E2), ubiquitin, ATP, and the target protein are incubated together. Following incubation, appearance of higher molecular weight target species on Coomassie stained gels or via immunoblotting is used as a qualitative read-out for ubiquitination. Because in vitro ubiquitination assays are conducted with purified recombinant proteins they are devoid of confounding elements present in the cellular milieu, allowing the establishment of a direct interaction between neosubstrate and Cereblon. In addition, the in vitro ubiquitination assay can be used to evaluate compound selectivity and specificity on a panel of compounds or to evaluate the role of individual amino acids in ternary complex formation through site-directed mutagenesis. In the methods section below, the in vitro ubiquitination assay is presented using a MBP-tagged recombinant substrate and ubiquitination is determined via immunoblotting.

286

1.5

2

Joel W. Thompson et al.

Cereblon Binding

Finally, a biochemical method is outlined to determine the affinity of CELMoDs for Cereblon. This Cereblon binding assay uses a TR-FRET-based method where CELMoD interacting with Cereblon displaces a Cy-5 labeled CELMoD from the tri-Trp pocket of Cereblon [19]. The Cereblon TR-FRET binding assay provides a sensitive and high-throughput screening format for the determination of IC50 values spanning low nM to mid μM. This assay is designed to handle either detailed dose-response analysis or can be scaled up for larger screening studies. Delineation of the IC50 value of CELMoD for Cereblon is an important parameter to parse how the elaboration of ligands affects CELMoD activity. It is known that for CELMoDs targeting the Cereblon neosubstrates Aiolos and Ikaros, increased Cereblon binding affinity can correlate with increased depth of degradation [20].

Materials

2.1 High-Throughput Cellular Substrate Degradation

1. Lenti-X 293T cells (Clonetech). 2. DMEM medium supplemented with 10% FBS. 3. DPBS, no calcium, no magnesium. 4. Trypsin-EDTA. 5. pMD2.G plasmid (lentiviral packaging vector). 6. psPAX2 plasmid (lentiviral envelope vector). 7. Plasmid encoding ePL-tagged protein of interest (lentiviral transfer vector with Puromycin selection marker). 8. Opti-MEM medium (Gibco). 9. Lipofectamine 2000 (Invitrogen). 10. 2 mL luer lock syringe. 11. 0.45 μm acetate syringe filter. 12. Polybrene. 13. Puromycin. 14. 10 cm and 15 cm tissue culture-treated dishes. 15. Bambanker freezing medium. 16. 384-well tissue-culture treated plates. 17. InCELL Detection Kit (DiscoverX). 18. DMSO.

2.2 Chemical and Genetic Approaches to Confirm CELMoD MOA

1. Human cell line. 2. Required media (culture medium, serum, growth factors, antibiotics, etc.) for culturing the desired human cell line. 3. Alt-R CRISPR-Cas9 tracrRNA, (IDT).

CELMoD Directed Proteolysis of CRBN Neosubstrates

287

4. Alt-R Cereblon crRNA ( AAAAUCCUGUUCUUCUCGAU ; IDT). 5. Alt-R S.p. HiFi Cas9 Nuclease V3 (IDT). 6. Nuclease-free Duplex Buffer (IDT). 7. 12-well tissue culture-treated plates. 8. Amaxa 4D-Nucleofector (Lonza). 9. Cell Line 4D-Nucleofector Kit S (Lonza). 10. Nucleocuvette strip (Lonza). 11. Bambanker freezing medium. 12. RIPA buffer. 13. 10
reducing agent (Invitrogen). 14. 4
LDS sample buffer (Invitrogen). 15. 4–12% Tris-Glycine SDS-PAGE gel. 16. Trans-blot Turbo Transfer System (Bio-Rad). 17. PBS Odyssey blocking buffer (LI-COR Biosciences). 18. Cereblon primary antibody. 19. Loading control primary antibody. 20. IRDye secondary antibodies (LI-COR Biosciences). 21. Odyssey CLx Imaging System (LI-COR Biosciences). 22. ImageStudio software (LI-COR Biosciences). 23. Primary antibody against protein of interest. 24. MLN4924 dissolved in DMSO. 25. MG132 dissolved in DMSO. 2.3 In Vitro Ubiquitination

1. QAH buffer: 20 mM HEPES pH 7.5, 150 mM NaCl, 10 mM MgCl2. 2. 5 μM recombinant human His6-Ubiquitin E1 enzyme (UBE1). 3. 25 μM recombinant human UbcH5a/UBE2D1 protein. 4. 30 μM recombinant human Ubiquitin protein (reconstituted in 500 μL of QAH buffer). 5. 25 μM recombinant human UbcH5b/UBE2D2 protein. 6. 0.5 M ATP in QAH Buffer. 7. 130 μM recombinant human Cereblon-DDB1 protein. 8. 28 μM recombinant human CUL4-Rbx1 protein. 9. Recombinant MBP-tagged substrate protein normalized to 1.4 mg/mL (30 μM). 10. DMSO (negative control). 11. 10 mM CELMoD (DMSO stock).

288

Joel W. Thompson et al.

12. 96-well PCR plate. 13. PCR plate seal. 14. 10
reducing agent (Invitrogen) 15. 2
SDS sample buffer (Invitrogen). 16. 4–12% Tris-Glycine SDS-PAGE gel. 17. iBlot 2 Dry Blotting System (ThermoFisher). 18. PBS Odyssey blocking buffer (LI-COR Biosciences). 19. MBP primary antibody. 20. IRDye secondary antibody (LI-COR Biosciences). 21. Odyssey CLx Imaging System (LI-COR Biosciences). 2.4

Cereblon Binding

1. 60 μM purified 6
His-Cereblon-DDB1 (Cereblon amino acids 1–442, DDB1 amino acids 1–1140). 2. 1.67 μM Eu-anti-His tag antibody. 3. 50 μM Cy5-conjugated CELMoD [20]. 4. FRET-assay buffer: 50 mM HEPES pH 7, 150 mM NaCl, and 0.005% Tween 20. 5. White 384-well assay plate. 6. DMSO. 7. 100 mM lenalidomide dissolved in DMSO.

3

Methods

3.1 High-Throughput Cellular Substrate Degradation 3.1.1 Generation of Cell Line Expressing ePLTagged Protein of Interest

1. Plate 3
105 Lenti-X 293T cells in 0.8 mL of antibiotic-free medium (DMEM with 10% FBS) per well of a 12-well plate and incubate cells ~24 h at 37  C/5% CO2. Include two additional wells of cells to be used for positive and negative controls. 2. Add 0.4 μg of pMD2.G plasmid, 0.4 μg of psPAX2 plasmid, and 0.8 μg lentiviral transfer plasmid encoding ePL-tagged protein of interest to 0.1 mL of Opti-MEM medium and incubate for 5 min at room temperature. A GFP encoding lentiviral transfer plasmid can be used as a positive control. As a negative control, a DNA mixture can be generated where the transfer plasmid is omitted. 3. Simultaneously, add 2.4 μL of Lipofectamine 2000 to 0.1 mL of Opti-MEM per condition and incubate for 5 min at room temperature. 4. Combine diluted plasmid DNA and Lipofectamine 2000. Mix DNA:Lipofectamine Opti-MEM mixtures and incubate for 20 min at room temperature.

CELMoD Directed Proteolysis of CRBN Neosubstrates

289

5. Add DNA:Lipofectamine Opti-MEM mixtures to previously plated Lenti-X 293T cells dropwise. 6. Incubate cells for ~16 h at 37  C/5% CO2. 7. Carefully remove medium from each well. Add 1.2 mL of fresh medium per well. 8. Incubate cells for ~30 h at 37  C/5% CO2. 9. While the Lenti-X 293T cells are producing lentivirus, fresh cells can simultaneously be plated for subsequent viral transductions. To this end, plate 5
105 Lenti-X 293T cells in 1.0 mL of medium per well of a 12-well plate and incubate at 37  C/5% CO2 until lentiviral supernatants are ready (see Note 1). Again, include two additional wells to be used as positive and negative control wells. 10. Approximately 30 h after adding fresh medium to transfected cells remove the lentivirus containing medium and pass it through a 0.45 μm syringe filter using a luer lock syringe. 11. Add 0.5 mL of individual viral supernatants to cells previously plated in step 9. The remaining viral supernatant can be stored away at 80  C. 12. Add polybrene at a final concentration of 5.0 μg/mL to each well of the 12-well plate. 13. Incubate cells for 24 h at 37  C/5% CO2. 14. Aspirate medium off of cells. Carefully wash cells with DPBS and trypsinize. Resuspend the cells and plate them in 15 mL of medium in a 10 cm dish. 15. Add puromycin to a final concentration of 1 μg/mL. 16. Incubate cells for ~72 h at 37  C/5% CO2. 17. Following 3 days of culturing under puromycin selection cells treated with the negative control viral supernatant should be dead. Cells treated with the GFP-positive control viral supernatant should be growing and healthy. GFP expression can be ascertained quickly using a florescence microscope or a flowcytometer. 18. Aspirate medium off dish. Carefully wash cells with DPBS and trypsinize. Resuspend the cells and plate them in 40 mL of medium containing 1 μg/mL puromycin in a 15-cm dish. 19. Incubate cells for ~72 h at 37  C/5% CO2. 20. Aspirate medium off of the dish. Carefully wash cells with DPBS and trypsinize. Plate 1/12 of the cells in 40 mL of medium containing 1 μg/mL puromycin in a 15-cm dish. Keep culturing these cells for future experiments. Set aside ~5
106 cells for pilot experiments (see Note 2). 21. Resuspend the remaining cells in Bambanker freezing medium and store away in multiple aliquots at 80  C.

290

Joel W. Thompson et al.

3.1.2 Optimization of Target-Based ePL Assay

1. From the cells obtained in step 20 from Subheading 3.1.1 create a 10 mL cell suspension in fresh medium containing 4
105 cells per mL. Make six 2-fold serial dilutions from this cell suspension. This will result in 7 different cell concentrations that when plated at a 25 μL volume will result in ~10
103, 5
103, 2.5
103, 1.25
103, 625, 312, and 156 cells per well. 2. In addition to the transduced cell lines repeat step 1 with parental Lenti-X 293T cells. This will be critical for calculating the signal to background ratio. 3. Plate cell dilutions into 12 wells of a 384-well plate using a 25 μL plating volume. 4. Incubate cells overnight at 37  C/5% CO2. 5. Remove 384-well plates from the incubator and leave at room temperature for 30 min. 6. Prepare InCELL Hunter Detection reagent according to manufacturer’s instructions (EA reagent, lysis buffer, and substrate reagent in a 1:1:4 ratio). 7. Add 25 μL of InCELL Hunter Detection reagent per well of a 384-well plate. Incubate plates 60 min at room temperature. 8. Read luminescence signal using a plate reader. 9. Plot luminescence values for a given cell line using a graphing software. Cells per well can be plotted on the x-axis and relative luminescence signal on the y-axis. Error bars for the luminescence signal can also be included by using the standard deviations for each cell dilution. At this point, it is critical to check that cell numbers correlate with the luminescent values. It is imperative that a cell plating value is chosen for future assays that ensures the assay is being conducted within the linear range. 10. Once the ideal cell number for a given ePL-tagged cell line has been determined, the signal to background value can also be calculated. To this end the luminescence value of the chosen cell number for the ePL-tagged stable cell line is divided by the luminescence value of the parental cell line at the same cell seeding density. Ideally a value 10 will be obtained but values as low as 5 can still give meaningful data.

3.1.3 Cellular Dose– Response Curve Assay of ePL-Tagged Protein of Interest

1. Dispense CELMoDs to be tested into a white 384-well tissue culture-treated plate using an acoustic liquid handler. Prepare dilutions based on a 25 μL assay volume in triplicate 10 point threefold serial dilutions starting with a 10 μM dose. 2. Include negative control wells containing only 0.1% DMSO to calculate 100% signal. Backfill all wells to a final DMSO concentration of 0.1% to ensure DMSO uniformity across wells.

CELMoD Directed Proteolysis of CRBN Neosubstrates

291

3. Wash, trypsinize, and count cells stably expressing ePL-tagged protein of interest. 4. Resuspend cells in fresh medium to give the proper concentration so when cells are plated at a 25 μL seeding volume, the assay will be conducted within the linear range as determined in step 9 from Subheading 3.1.2. 5. Dispense 25 μL of cells expressing ePL-tagged protein of interest per well. Seed cells into the 384-well plate that was prespotted with compounds in step 1. 6. Incubate cells overnight at 37  C/5% CO2. 7. Remove 384-well plates from the incubator and leave at room temperature for 30 min. 8. Prepare InCELL Hunter Detection reagent according to manufacturer’s instructions (EA reagent, lysis buffer, and substrate reagent in a 1:1:4 ratio). 9. Add 25 μL of InCELL Hunter Detection reagent per well of a 384-well plate. Incubate plates 60 min at room temperature. 10. Read luminescence signal using a plate reader. 11. Normalize all luminescence values to the DMSO control wells. The average value of the DMSO control wells should be set to equal 100% of the relative ePL-tagged target protein levels. Plot luminescence values using a graphing software. CELMoD concentration can be plotted on the x-axis and the corresponding relative protein of interest levels on the y-axis. Use the graphing software to determine the DC50 value (the half-maximum effective concentration) of a CELMoD for the degradation of the ePL-tagged substrate. Most software will use a four-parameter logistic model (sigmoidal doseresponse model) (FIT ¼ (A + {(B  A)/1 + [(C/x)D]})), where C is the inflection point (DC50), D is the correlation coefficient, and A and B are the low and high limits of the fit, respectively) to calculate the DC50. Calculate the Ymin by determining the lowest percentage of target protein remaining following CELMoD treatment (Fig. 1). 3.2 Chemical and Genetic Approaches to Confirm CELMoD MOA 3.2.1 Generation of Cereblon Knockout Cells

1. Resuspend tracrRNA and crRNA to 200 μM each in NucleaseFree Duplex Buffer. 2. In microfuge tubes, mix crRNA 1:1 with tracrRNA. A crRNA targeting HPRT can be used as a positive control. As a negative control, a nontargeting crRNA can be employed. 3. Heat crRNA/tracrRNA mixture at 95  C for 5 min and cool to room temperature on bench top. 4. Mix 2 μL crRNA/tracrRNA (sgRNA) mixture with 1 μL Cas9 (10 μg/μL) to make ribonucleoprotein (RNP) mixture. 5. Incubate RNP mixture at room temperature for 10 min.

292

Joel W. Thompson et al.

Fig. 1 High-throughput Cellular Substrate Degradation Assay. A human cell line expressing an ePL-tagged substrate was treated for 6 h. with the indicated doses of CELMoD A, CELMoD B, or CELMoD C. Data is presented as mean  SD of 3 replicates. DC50 is the concentration of a drug that gives half-maximal degradation of the substrate and Ymin is the lowest percentage of substrate remaining following drug treatment

6. Put 1 mL of medium per sample in one well of a 12-well plate and place plate in incubator to prewarm media. 7. Make working nucleofection solution by adding 16.4 μL Nucleofector solution to 3.6 μL Supplement 1 per sample (see Note 3). 8. Pellet cells (4
105/sample) and wash with PBS. 9. Resuspend cells in 20 μL of working nucleofection solution. 10. Combine 3 μL of RNP mixture and 20 μL of resuspended cells in nucleofection solution and place into one well of a nucleocuvette strip per sample. 11. Electroporate cells using an Amaxa 4D-Nucleofector (see Note 4). 12. Transfer electroporated cells to a well with prewarmed media in a 12-well plate from Step 6. 13. Incubated cells at 37  C/5% CO2. 14. Expand cells until there is a sufficient amount to freeze down (~2
107). 15. Resuspend cells in Bambanker freezing medium and store away in multiple aliquots at 80  C. 16. To analyze Cereblon knockout efficiency, resuspend 1
106 of the parental cells and 1
106 cells that were electroporated with Cereblon sgRNA in 75 μL of ice cold RIPA buffer per sample (see Note 5). 17. Centrifuge lysates at 15,000
g for 10 min at 4  C.

CELMoD Directed Proteolysis of CRBN Neosubstrates

293

18. Combine 48.75 μL of clarified lysate with 7.5 μL of 10
reducing agent and 18.75 μL of 4
LDS sample buffer. 19. Heat samples at 100  C for 8 min. 20. Load 15 μL of each sample onto on a 4–12% Tris-Glycine SDS-PAGE gel. 21. Run gel at 180 V until dye front reaches the bottom of the gel. 22. Transfer proteins from gel to a nitrocellulose membrane using the trans-blot turbo transfer system. 23. Place membrane in plastic container and block for 1 h in PBS Odyssey blocking buffer. 24. Add Cereblon and loading control primary antibodies diluted in PBST to membrane. Incubate membrane at 4  C with agitation for 16 h. 25. Wash membrane five times for 5 min with PBST. 26. Incubate membrane with the appropriate anti-IgG secondary IRDye antibodies (1:20,000 dilution in PBST) for 1 h at room temperature. 27. Wash membrane five times for 5 min with PBST. 28. Capture immunoblot images using an Odyssey imaging system. 29. Quantitate the relative intensity of Cereblon expression in the parental and Cereblon KO cell lines using ImageStudio software. If Cereblon expression in the Cereblon KO cell line is reduced to less than 10% of the parental cell line, proceed with additional experiments (see Note 6). 3.2.2 Confirmation of Cereblon- and Proteasome-Dependent Regulation of Protein of Interest Via CELMoD Treatment

1. Prespot all compounds into a 12-well plate based on a 1 mL assay volume. For both the parental and Cereblon KO cell lines there will be six conditions—(1) vehicle control (DMSO), (2) CELMoD alone, (3) MG132 alone, (4) MG132 + CELMoD, (5) MLN4924 alone, and (6) MLN4924 + CELMoD. Prespot CELMoDs at a final concentration of 10 μM, MLN4924 at a final concentration of 1 μM, and MG132 at a final concentration of 10 μM. Be sure to normalize every well to ensure it contains the same percentage of DMSO. 2. Plate 2
106 parental and Cereblon KO cells per condition in 1.0 mL of media per well of a 12-well plate prespotted with compounds. 3. Incubated cells for ~6 h at 37  C/5% CO2. 4. Lyse cells in 60 μL of ice cold RIPA buffer per sample. 5. To analyze samples via immunoblotting refer to steps 17–28 of Subheading 3.2.1. Make sure to analyze the relative expression of the protein of interest in addition to Cereblon and the loading control (see Note 7).

294

Joel W. Thompson et al.

Fig. 2 Confirmation of Cereblon-dependent CELMoD MOA. Immunoblot analysis of endogenous substrate in parental or Cereblon (CRBN) knockout (KO) human cells incubated for 6 h with DSMO, 10 μM CELMoD x, 10 μM CELMoD y, or 10 μM CELMoD z. CELMoD x, CELMoD y, and CELMoD z appear to be degrading substrate via a Cereblon-dependent mechanism as substrate protein levels decrease in parental cells but remain unchanged in Cereblon KO cells following drug treatment

6. After capturing the immunoblot images, if the target of interest’s protein level is unchanged following CELMoD treatment in the Cereblon KO cells but the proteins levels are decreased when treated with CELMoD in parental cells, this will suggest the CELMoD tested is acting via a Cereblon-dependent MOA to degrade the protein of interest (Fig. 2). Likewise, if the target of interest’s protein level is unchanged following CELMoD treatment in cells cotreated with MG132 or MLN4924 but the level decreases with CELMoD alone, this will suggest the CELMoD tested is acting via a proteasomal- and Cereblondependent mechanism to degrade the protein of interest. 3.3 In Vitro Ubiquitination

Thaw all frozen reagents on ice. Keep reagents on ice except where indicated otherwise. 1. Prepare E1/E2 master mix. Each reaction contains 1.5 μL UBE1, 3 μL of UBE2D1, 0.3 μL of UBE2D2, 1.5 μL of Ubiquitin, and 3.7 μL of ATP. An excess is made in order to account for any dead volume. 2. Incubate E1/E2 master mix at room temperature for 30 min before combining with other reaction components (see Note 8). 3. Prepare CELMoD (50 μM) and DMSO dilutions by adding 1 μL of 10 mM CELMoD to 199 μL of QAH buffer, mixing well. Add 1 μL of DMSO to 199 μL of QAH buffer, mixing well. 4. Prepare and aliquot DMSO/E3 mix (see Note 9). Each reaction contains 4 μL of diluted DMSO, 0.5 μL of CereblonDDB1 protein, and 0.5 μL of CUL4-Rbx1 protein. An excess

CELMoD Directed Proteolysis of CRBN Neosubstrates

295

is made in order to account for any dead volume. Mix the final solution gently with a pipette. Aliquot 5 μL of DMSO/E3 mix into a 96-well PCR plate. 5. Prepare and aliquot CELMoD/E3 Mix. Each reaction contains 4 μL of 50 μM CELMoD, 0.5 μL of Cereblon-DDB1 protein, and 0.5 μL of CUL4-Rbx1protein. An excess is made in order to account for any dead volume. Mix the final solution gently with a pipette. Prepare a separate CELMoD/E3 mix for each CELMoD. Aliquot 5 μL of each CELMoD/E3 mix to the PCR plate. 6. Add 5 μL substrate to each well of aliquoted CELMoD/E3 and DMSO/E3 mixes in the PCR plate. 7. Add 10 μL of E1/E2 master mix to each reaction well in the PCR plate. Seal PCR plate and spin in centrifuge at 100
g for 30 s. 8. Incubate reaction. Add PCR plate to PCR machine and heat the ubiquitination reactions for 2 h at 30  C. 9. Combine 3 μL of the in vitro ubiquitination reaction mixture with 5 μL of QAH, 2 μL of 10
reducing agent, and 10 μL of 2
SDS sample buffer. 10. Heat samples at 100  C for 8 min. 11. Load 20 μL of each sample and 3 μL protein standard onto on a 4–12% Tris-Glycine SDS-PAGE gel. 12. Run gel at 180 V until dye front reaches the bottom of the gel (see Note 10). 13. Transfer proteins from gel to a nitrocellulose membrane using the iBlot 2 Dry Blotting system. 14. Place membrane in plastic container and block for 1 h in PBS Odyssey blocking buffer. 15. Add MBP primary antibody diluted in PBST to membrane. Incubate membrane at 4  C with agitation for 16 h. 16. Wash membrane five times for 5 min with PBST. 17. Incubate membrane with the appropriate anti-IgG secondary IRDye antibody (1:30,000 dilution in PBST) for 1 h at room temperature. 18. Wash membrane five times for 5 min with PBST. 19. Capture immunoblot images using an Odyssey imaging system. 20. Qualitative analysis of ubiquitination reactions can be performed by comparing the relative laddering pattern of substrate across CEMoD treatments to the DMSO control reaction (see Note 11) (Fig. 3).

296

Joel W. Thompson et al.

Fig. 3 CELMoD-mediated In Vitro Ubiquitination Assay. MBP-tagged substrate was incubated with CRL4Cereblon and DMSO (vehicle control), or three different CELMoDs. Ubiquitination reactions were separated by SDS-PAGE followed by anti-MBP immunoblotting. Treatment with CELMoD II leads to the greatest level of substrate ubiquitination followed by CELMoD III and CELMoD I 3.4

Cereblon Binding

1. Add 11.96 mL of FRET-assay buffer into a 15 mL conical tube. 2. Add 11 μL of 60 μM 6
His-Cereblon-DDB1 to assay buffer. 3. Add 22 μL of 1.67 μM Eu anti-His tag antibody to assay mixture. 4. Add 7.2 μL of 50 μM Cy5-conjugated CELMoD to assay mixture. 5. Invert conical tube several times to ensure components are mixed. 6. Dispense CELMoDs to be tested into a white 384-well assay plate using an acoustic liquid handler. Prepare dilutions based on a 30 μL assay volume in duplicate 10 point threefold serial dilutions starting with a 100 μM dose. 7. Include negative control wells containing only 0.1% DMSO to demonstrate 100% signal. Positive controls wells to demonstrate 0% signal can be obtained by dispensing lenalidomide at a final assay concentration of 100 μM. Backfill all wells to a final DMSO concentration of 0.1% to ensure DMSO uniformity across wells. 8. Dispense 30 μL of assay mix into each testing well of the 384-well plate to initiate the reaction. 9. Place plates on a shaker for 10 s to mix. 10. Give plates a brief spin in a centrifuge to ensure the assay mixture is in the bottom of the wells in the assay plate. 11. Incubate plates at room temperature for 10 min (see Note 12). 12. Read plates on a FRET-compatible plate reader with excitation at 340 nm and monitor emissions at 615 nm (non-FRET) and 665 nm (FRET) using a 700 μs TR-FRET delay. 13. Calculate the FRET/non-FRET emission ratio (ratio of 665 nm/615 nm) for each compound concentration to

CELMoD Directed Proteolysis of CRBN Neosubstrates

297

Fig. 4 Cereblon Binding Assay. Determination of the relative Cereblon binding affinities for CELMoDs by a TR-FRET displacement assay. IC50 values for the compounds tested were 25 nM for CELMoD 1, 177 nM for CELMoD 2, 8.2 μM for CELMoD 3, and 14 μM for CELMoD 4

determine the FRET efficiency. Normalize all FRET efficiency values to the DMSO control wells. The average value of the DMSO control wells should be set to equal 100. 14. Plot normalized FRET efficiency values against compound concentration and determine the relative binding affinity (IC50) using a sigmoidal 4-parameter fit as shown below (see Note 13) (Fig. 4). y ¼dþa

d 1þ


x b c

The 4 estimated parameters consist of the following: a ¼ the minimum value that can be obtained (fill inhibition/no activity). d ¼ the maximum value that can be obtained (100% activity). c ¼ IC50 (i.e., the point on the S-shaped curve halfway between a and d). b ¼ Hill’s slope of the curve (i.e., this is related to the steepness of the curve at point c).

298

4

Joel W. Thompson et al.

Notes 1. At this stage of the procedure, Lenti-X 293T cells do not have to be chosen to generate the ePL-tagged stable cell line. Any immortalized human cell line can be selected provided it is amenable to lentiviral transduction and has robust active levels of Cereblon. 2. 1
106 cells from step 20 Subheading 3.1.1 can be analyzed by SDS-PAGE/Immunoblotting for confirmation of expression of an ePL-tagged protein with the correct predicted molecular weight. 3. Go to Lonza.com to determine which Nucleofector kit solution is optimal for the human cell line selected for making Cereblon knockout cells. 4. Go to Lonza.com to determine which Nucleofector program is optimal based on the Nucleofector kit solution chosen for the human cell line being electroporated. 5. As an alternative or orthogonal approach to immunoblotting, TIDE analysis can be used to determine Cereblon knockout efficiency. 6. If Cereblon knockout efficiency is poor, Cereblon knockout cells can be plated as single cells or in limited dilutions to generate a cell line with robust Cereblon knockout efficiency. 7. If a primary antibody against the protein of interest is not available, a cell line stably expressing an epitope-tagged transgene encoding the protein of interest can be used. 8. Beginning the protocol by making the E1/E2/Ubiquitination master mix precharges the E2 with ubiquitin and produces a more efficient ubiquitination reaction. 9. Running a DMSO negative control reaction establishes the degree of compound independent ubiquitination for each substrate. In addition, running a substrate alone, lane (not shown in Fig. 3) provides a reference for substrate location on the gel, a control for substrate load, and a reference for evaluating substrate depletion. 10. Using chilled SDS running buffer and lower voltage will reduce gel smiling and produce higher quality immunoblot images. 11. If substrate ubiquitination is robust it may be readily discerned via Coomassie staining, and immunoblotting analysis may not be necessary. Typically, a 10–20 μL load of the ubiquitination reaction is used for Coomassie analysis.

CELMoD Directed Proteolysis of CRBN Neosubstrates

299

12. Plate is typically read 10 min after incubation. However, the FRET signal is stable for ~30–45 min before it begins to slowly decrease. 13. Alternatively, graphing software such as XLfit, Prism, etc. can be used to determine IC50 values.

Acknowledgments Sincerest appreciation to all Celgene/BMS employees who have contributed to the advancement of CELMoD research. References 1. Raina K, Crews CM (2017) Targeted protein knockdown using small molecule degraders. Curr Opin Chem Biol 39:46–53. https://doi. org/10.1016/j.cbpa.2017.05.016 2. Huang X, Dixit VM (2016) Drugging the undruggables: exploring the ubiquitin system for drug development. Cell Res 26 (4):484–498. https://doi.org/10.1038/cr. 2016.31 3. Schapira M, Calabrese MF, Bullock AN, Crews CM (2019) Targeted protein degradation: expanding the toolbox. Nat Rev Drug Discov 18(12):949–963. https://doi.org/10.1038/ s41573-019-0047-y 4. Chamberlain PP, Hamann LG (2019) Development of targeted protein degradation therapeutics. Nat Chem Biol 15(10):937–944. https://doi.org/10.1038/s41589-019-0362y 5. Ito T, Ando H, Suzuki T, Ogura T, Hotta K, Imamura Y, Yamaguchi Y, Handa H (2010) Identification of a primary target of thalidomide teratogenicity. Science 327 (5971):1345–1350. https://doi.org/10. 1126/science.1177319 6. Chamberlain PP, Cathers BE (2019) Cereblon modulators: low molecular weight inducers of protein degradation. Drug Discov Today Technol 31:29–34. https://doi.org/10.1016/j. ddtec.2019.02.004 7. Chamberlain PP, D’Agostino LA, Ellis JM, Hansen JD, Matyskiela ME, McDonald JJ, Riggs JR, Hamann LG (2019) Evolution of Cereblon-mediated protein degradation as a therapeutic modality. ACS Med Chem Lett 10 (12):1592–1602. https://doi.org/10.1021/ acsmedchemlett.9b00425 8. Gandhi AK, Kang J, Havens CG, Conklin T, Ning Y, Wu L, Ito T, Ando H, Waldman MF, Thakurta A, Klippel A, Handa H, Daniel TO,

Schafer PH, Chopra R (2014) Immunomodulatory agents lenalidomide and pomalidomide co-stimulate T cells by inducing degradation of T cell repressors Ikaros and Aiolos via modulation of the E3 ubiquitin ligase complex CRL4 (CRBN.). Br J Haematol 164(6):811–821. https://doi.org/10.1111/bjh.12708 9. Kronke J, Udeshi ND, Narla A, Grauman P, Hurst SN, McConkey M, Svinkina T, Heckl D, Comer E, Li X, Ciarlo C, Hartman E, Munshi N, Schenone M, Schreiber SL, Carr SA, Ebert BL (2014) Lenalidomide causes selective degradation of IKZF1 and IKZF3 in multiple myeloma cells. Science 343 (6168):301–305. https://doi.org/10.1126/ science.1244851 10. Lu G, Middleton RE, Sun H, Naniong M, Ott CJ, Mitsiades CS, Wong KK, Bradner JE, Kaelin WG Jr (2014) The myeloma drug lenalidomide promotes the cereblon-dependent destruction of Ikaros proteins. Science 343 (6168):305–309. https://doi.org/10.1126/ science.1244917 11. An J, Ponthier CM, Sack R, Seebacher J, Stadler MB, Donovan KA, Fischer ES (2017) pSILAC mass spectrometry reveals ZFP91 as IMiD-dependent substrate of the CRL4 (CRBN) ubiquitin ligase. Nat Commun 8:15398. https://doi.org/10.1038/ ncomms15398 12. Donovan KA, An J, Nowak RP, Yuan JC, Fink EC, Berry BC, Ebert BL, Fischer ES (2018) Thalidomide promotes degradation of SALL4, a transcription factor implicated in Duane radial ray syndrome. Elife 7. https://doi.org/ 10.7554/eLife.38430 13. Kronke J, Fink EC, Hollenbach PW, MacBeth KJ, Hurst SN, Udeshi ND, Chamberlain PP, Mani DR, Man HW, Gandhi AK, Svinkina T, Schneider RK, McConkey M, Jaras M,

300

Joel W. Thompson et al.

Griffiths E, Wetzler M, Bullinger L, Cathers BE, Carr SA, Chopra R, Ebert BL (2015) Lenalidomide induces ubiquitination and degradation of CK1alpha in del(5q) MDS. Nature 523(7559):183–188. https://doi.org/10. 1038/nature14610 14. Matyskiela ME, Couto S, Zheng X, Lu G, Hui J, Stamp K, Drew C, Ren Y, Wang M, Carpenter A, Lee CW, Clayton T, Fang W, Lu CC, Riley M, Abdubek P, Blease K, Hartke J, Kumar G, Vessey R, Rolfe M, Hamann LG, Chamberlain PP (2018) SALL4 mediates teratogenicity as a thalidomide-dependent cereblon substrate. Nat Chem Biol 14 (10):981–987. https://doi.org/10.1038/ s41589-018-0129-x 15. Matyskiela ME, Lu G, Ito T, Pagarigan B, Lu CC, Miller K, Fang W, Wang NY, Nguyen D, Houston J, Carmel G, Tran T, Riley M, Nosaka L, Lander GC, Gaidarova S, Xu S, Ruchelman AL, Handa H, Carmichael J, Daniel TO, Cathers BE, Lopez-Girona A, Chamberlain PP (2016) A novel cereblon modulator recruits GSPT1 to the CRL4(CRBN) ubiquitin ligase. Nature 535(7611):252–257. https://doi.org/10.1038/nature18611 16. Sievers QL, Petzold G, Bunker RD, Renneville A, Slabicki M, Liddicoat BJ, Abdulrahman W, Mikkelsen T, Ebert BL, Thoma NH (2018) Defining the human C2H2 zinc finger degrome targeted by thalidomide analogs through CRBN. Science 362 (6414). https://doi.org/10.1126/science. aat0572

17. Schwinn MK, Machleidt T, Zimmerman K, Eggers CT, Dixon AS, Hurst R, Hall MP, Encell LP, Binkowski BF, Wood KV (2018) CRISPR-mediated tagging of endogenous proteins with a luminescent peptide. ACS Chem Biol 13(2):467–474. https://doi.org/ 10.1021/acschembio.7b00549 18. Olson KR, Eglen RM (2007) Beta galactosidase complementation: a cell-based luminescent assay platform for drug discovery. Assay Drug Dev Technol 5(1):137–144. https://doi. org/10.1089/adt.2006.052 19. Chamberlain PP, Lopez-Girona A, Miller K, Carmel G, Pagarigan B, Chie-Leon B, Rychak E, Corral LG, Ren YJ, Wang M, Riley M, Delker SL, Ito T, Ando H, Mori T, Hirano Y, Handa H, Hakoshima T, Daniel TO, Cathers BE (2014) Structure of the human Cereblon-DDB1-lenalidomide complex reveals basis for responsiveness to thalidomide analogs. Nat Struct Mol Biol 21(9):803–809. https://doi.org/10.1038/nsmb.2874 20. Matyskiela ME, Zhang W, Man HW, Muller G, Khambatta G, Baculi F, Hickman M, LeBrun L, Pagarigan B, Carmel G, Lu CC, Lu G, Riley M, Satoh Y, Schafer P, Daniel T, Carmichael J, Cathers BE, Chamberlain PP (2018) A Cereblon modulator (CC-220) with improved degradation of Ikaros and Aiolos. J Med Chem 61(2):535–542. https://doi.org/ 10.1021/acs.jmedchem.6b01921

Chapter 16 Global Proteome Profiling to Assess Changes in Protein Abundance Using Isobaric Labeling and Liquid Chromatography-Tandem Mass Spectrometry Anthony P. Possemato, Kathryn Abell, and Matthew P. Stokes Abstract Protein degradation is a critical component of all facets of cell biology, and recently methods have been developed to make use of targeted protein degradation as both an investigative tool and a potential therapeutic avenue. Mass spectrometry-based proteomic studies have allowed detailed characterization of changes in protein level and the biology underlying growth, development, and disease. Current methods and instrumentation allow identification and quantitative analysis of thousands of proteins in a single assay. The method described here involves cell lysis and digestion to peptides, labeling peptides with isobaric tagging TMT reagents, basic reversed phase fractionation, and liquid chromatography-tandem mass spectrometry analysis of the enriched peptides. Key words Mass spectrometry, LC-MS/MS, Proteomics, Isobaric Tag, TMT, Proteome

1

Introduction Changes in protein levels and protein degradation affect nearly all facets of biology, including normal growth, development, and disease [1–5]. In addition to normal fluctuations in protein levels, strategies for targeted protein degradation have been developed that can be used to answer research questions or as therapeutic agents [6–8]. Methods that allow identification of proteins that change in abundance between samples and quantify the magnitude of those changes are powerful tools for better understanding protein turnover and degradation. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is one such tool that has been used to profile many biological systems to find changes in protein abundance on a global scale [9–13]. In these mass spectrometry-based studies, there are a number of options for quantification, including label-free analysis of relative peptide abundance, labeling methods such as SILAC or reductive

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_16, © Springer Science+Business Media, LLC, part of Springer Nature 2021

301

302

Anthony P. Possemato et al.

dimethylation, and isobaric tagging [14–21]. Label-free quantification is the most direct measure of relative peptide/protein abundance; however, depth of coverage in label-free studies can be difficult, as fractionation of samples is not recommended. Labelfree is best suited for experimental systems where some enrichment is performed on the samples prior to LC-MS/MS analysis, such as PTMScan profiling of post-translationally modified peptides [20, 22, 23]. Label-free quantification also relies on reproducible chromatography between samples and consistent performance of the mass spectrometer across many runs. SILAC and reductive dimethylation allow analysis of multiple samples in a single LCMS run; however, both have the disadvantage of the instrument spending time (duty cycle) identifying the heavy and light forms of the same peptide, reducing the number of peptides/proteins identified. SILAC has the further disadvantage of requiring growth through many doublings in specialized media, and is not easily amenable to work in tissues. Both methods are also limited in number of samples that can be multiplexed to 2 or 3. Isobaric tagging is a labeling method that makes use of 4, 6, 8, 10, 11, or even 16 different tags which all have the same mass when added to peptides. In addition to the higher plex afforded by these reagents compared to SILAC, the fact that the tags have the same mass means that the instrument only spends time identifying one “version” of the peptide in the MS1 channel. Once the peptides go to MS2, the fragment of the tags that breaks off (reporter ions) each have a different mass, allowing quantification of the relative amount of each peptide across the multiplexed samples. One major drawback of isobaric tagging studies is that because the reporter ions are physically separated from the peptide to be quantified, there can be contamination of one peptide’s signal with reporter ions from another peptide. This can cause ratio compression, decreasing the accuracy of quantification. A necessary workaround for this issue is MS3-based quantification, in which an additional round of fragmentation is used to further increase selectivity and quantitative accuracy [15, 17, 24–26]. In this method, cells, tissues, biofluids, or other biological materials are lysed and digested with LysC and trypsin, generating peptides. These peptides are purified using reversed-phase (hydrophobic) solid-phase extraction. Purified peptides are then labeled with the TMT isobaric tagging reagents, and efficiency of labeling is checked. Labeled peptides are fractionated using high pH reversed phase chromatography over a C18 column, and fractions are concatenated nonsequentially to maximize peptide identification rate across all samples run on the mass spectrometer (Fig. 1). The concatenated fractions are then run in LC-MS/MS using a multinotch MS3 method where peptides are identified and reporter ion intensities measured. This peptide-level data is then collapsed down

Total Proteome Mass Spectrometry Analysis

303

Fig. 1 bRP fractionation and concatenation Scheme. 96 bRP fractions are collected and combined nonsequentially to 12 fractions to be run in LC-MS/MS. This ensures that all fractions that go to the mass spectrometer have roughly equal peptide content and maximize LCMS efficiency

to the protein level for a global view of protein abundance and changes in protein abundance across samples.

2

Materials 1. 1
Phosphate-buffered Technology, 9808).

saline

(PBS)

(Cell

Signaling

2. 2 mL microcentrifuge tubes and 15 mL conical tubes. 3. Centrifuges capable of handling 15 mL tubes and 2 mL tubes. 4. Urea, Ultrapure, PTMScan Qualified (CST 60055). 5. 200 mM HEPES, pH 8.0 (CST 44686). 6. Phosphatase Inhibitor Cocktail (CST 5870). 7. Sonicator with microtip. 8. Polytron electronic homogenizer. 9. Protein Quantitation Colorimetric Assay Kit, such as Pierce Coomassie Plus (Bradford) Assay Kit (Thermo 23236). 10. Dithiothreitol (DTT) (Cell Signaling Technology 7016). 11. Iodoacetamide, PTMScan Qualified (Cell Signaling Technology 88931). 12. PTMScan Lys-C Protease (Cell Signaling Technology 39003). 13. PTMScan Trypsin, TPCK-Treated (Cell Signaling Technology 56296). 14. 1 mM Hydrochloric acid (HCl).

304

Anthony P. Possemato et al.

15. End-over-end rotator. 16. Water, Burdick & Jackson LCMS-grade (CST 27732). 17. Trifluoroacetic Acid (TFA), sequencing grade (Thermo 28904). 18. Sep-Pak C18 1 cm3 Vac Cartridge, 50 mg Sorbent per Cartridge (Waters WAT054955). 19. Acetonitrile (ACN), PTMScan Qualified (Cell Signaling Technology, 95031). 20. Lyophilizer. 21. Vacuum concentrator (Speed-Vac). 22. Pierce C18 Spin Tips (Thermo 84850). 23. Empore C18 disks (CDS Analytical 2215 or 2315). 24. 18 gauge blunt tip needle. 25. HEPES, sodium salt (Sigma H7006). 26. TMT10plex™ Isobaric Label Reagent Set (Thermo 90110). 27. 50% Hydroxylamine (Thermo 90115). 28. 28% Ammonium hydroxide solution (Sigma 338818). 29. Formic Acid, LCMS grade (Fisher A117-50 or similar quality for LCMS analysis). 30. Agilent 1260 Infinity II HPLC (G7111B) with integrated fraction collector. 31. C18 resin or C18 column such as Accucore C18 2.6 μm, 150 Å (Thermo). 2.1 Lysis and Digestion to Peptides

1. Urea Lysis Buffer: 20 mM HEPES pH 8.0, 9 M urea, 2
phosphatase inhibitor cocktail. Add 10 mL of 200 mM HEPES to 54.05 g urea and 2 mL of phosphatase inhibitor cocktail. Add reverse osmosis deionized (RODI) or equivalent water to 100 mL total. Mix well until the urea goes completely into solution. 2. 20 mM HEPES pH 8.0: Dilute 10 mL 200 mM HEPES stock solution in 90 mL RODI or equivalent water. 3. Dithiothreitol (DTT): Make 1.25 M stock. Resuspend one tube containing 192.8 mg with 1 mL water. Divide into 25 μL aliquots. Store at 20  C for up to 1 year. Thaw one aliquot for each experiment. 4. Iodoacetamide solution: Make 100 mM stock. Weigh out 19 mg of iodoacetamide and cover the tube with foil to protect it from light. Dissolve the powder in water to a final volume of 1 mL immediately before use. The iodoacetamide solution should be prepared fresh prior to each experiment.

Total Proteome Mass Spectrometry Analysis

305

5. LysC Protease: Make a 5 mg/mL stock solution. Resuspend 1.2 mg LysC in 240 μL of 20 mM HEPES pH 8.0, aliquot for single use and store at 80  C. 6. Trypsin-TPCK: Make a 1 mg/mL solution. Dissolve 20 mg trypsin in 20 mL of 1 mM HCl prior to use. Aliquot and store at 80  C. 2.2 Sep-Pak® C18 Purification of Lysate Peptides

1. 20% Trifluoroacetic acid (TFA): Add 20 mL TFA to 80 mL LCMS grade water. Store at room temperature. 2. Sep-Pak Wetting solution: 100% acetonitrile (ACN). 3. Sep-Pak Wash Solution A: 0.1% TFA. Add 1 mL 20% TFA to 199 mL LCMS grade water. Store at room temperature. 4. Sep-Pak Wash Solution B: 0.1% TFA/5% ACN. Add 1 mL 20% TFA to 189 mL LCMS grade water, mix in 10 mL 100% ACN. Store at room temperature in a well-sealed tube/bottle to prevent evaporation of ACN. 5. Sep-Pak Elution Solution: 0.1% TFA/40% ACN. Add 1 mL 20% TFA to 119 mL LCMS grade water, mix in 80 mL 100% ACN. Store at room temperature in a well-sealed tube/bottle to prevent evaporation of ACN.

2.3

TMT Labeling

2.4 C18 Sample Desalting

1. 200 mM Na-HEPES pH 8.5. Add 26.03 mg Na-HEPES to 80 mL RODI or equivalent water. pH to 8.5 with 5 M NaOH, bring volume to 100 mL with water. Store at room temperature or 4  C. 1. C18 tip Wetting Solution: 50% ACN, 0.1% TFA. Add 5 μL 20% TFA to 495 μL LCMS grade water, mix with 500 μL ACN. Store at room temperature in a well- sealed tube/bottle to prevent evaporation of ACN. 2. C18 tip Equilibrate & Wash Solution: 0.1% TFA. Add 5 μL 20% TFA to 995 μL LCMS grade water. Store at room temperature. 3. C18 tip Elution Solution: 40% ACN, 0.1% TFA. Add 5 μL 20% TFA to 595 μL LCMS grade water, mix with 400 μL ACN. Store at room temperature in a well- sealed tube/bottle to prevent evaporation of ACN.

2.5 Basic Reversed Phase (bRP) Fractionation [11]

1. 10% formic acid. Add 100 mL formic acid to 900 mL LCMS grade water. 2. 180 mM ammonium formate, pH 10. Add 25 mL of 28% ammonium hydroxide to ~500 mL of LCMS grade water, then add ~45 mL of 10% formic acid to pH 10.0, bring final volume to 1 L with LCMS grade water. Check the pH immediately before use to prepare bRP solvents.

306

Anthony P. Possemato et al.

3. bRP Solvent A: 5% acetonitrile in 10 mM ammonium formate pH 10. Add 55.5 mL 180 mM ammonium formate pH 10.0 to 894.5 mL LCMS grade water, then add 50 mL ACN. 4. bRP Solvent B: 80% acetonitrile in 10 mM ammonium formate pH 10. Add 55.5 mL 180 mM ammonium formate pH 10.0 to 144.5 mL LCMS grade water, then add 800 mL ACN. 2.6 LC-MS/MS Analysis

1. LCMS Solvent A: 3% acetonitrile in 0.12% formic acid. Add 1.2 mL formic acid to 968.8 mL LCMS grade water, mix with 30 mL ACN. Store at room temperature in a well-sealed bottle to prevent evaporation of ACN. 2. LCMS Solvent B: 80% acetonitrile in 0.12% formic acid. Add 1.2 mL formic acid to 198.8 mL LCMS grade water, mix with 800 mL ACN. Store at room temperature in a well-sealed bottle to prevent evaporation of ACN. 3. 50 cm
100 μm capillary column packed with C18 reversedphase resin.

3

Methods Carry out all steps at room temperature unless otherwise noted.

3.1 Lysis and Digestion to Peptides

1. 100 μg total protein is recommended as the starting amount for all samples, though the protocol can be adapted for processing smaller sample amounts. 100 μg corresponds to ~1E6–4E6 cells or 1–4 mg wet tissue weight. 2. For tissue samples, resuspend flash frozen tissue in Urea Lysis Buffer, for suspension cells resuspend PBS-washed cell pellet in Urea Lysis Buffer, and for adherent cells scrape PBS-washed cells off plates using Urea Lysis Buffer. 3. For tissue samples homogenize using electronic homogenizer such as a PolyTron 2
30 s. 4. Sonicate samples at 5–15 W output power 3
15 s each, cooling on ice for 1 min between each burst (see Note 1). 5. Centrifuge samples 15 min at room temperature at 10,000
g (see Note 2). 6. Collect supernatant from centrifugation as cleared protein lysate. 7. Perform protein assay using Pierce Coomassie Plus (Bradford) Assay Kit and use 100 μg total protein per sample for following steps (see Note 3). 8. Reduce samples with 1/278 volume of 1.25 M DTT to the cleared protein lysate (e.g., 3.6 μL of 1.25 M DTT for 1 mL of protein extract), mix well and incubate at room temperature for 60 min (see Note 4).

Total Proteome Mass Spectrometry Analysis

307

9. Alkylate samples by adding 1:10 volume of iodoacetamide solution, mixing well, and incubating for 15 min in the dark. 10. Digest samples for 2 h at 37  C with 1:100 (w/w) dilution of LysC. Dilute 5 mg/mL stock 1:10 with 20 mM HEPES pH 8.0 to 0.5 mg/mL and add 2 μL (1 μg enzyme) of this diluted stock to digest 100 μg protein. 11. Dilute samples 1:4 with 20 mM HEPES pH 8.0 and digest overnight at room temperature with gentle rocking with 1:100 (v/v) trypsin (see Note 5). 3.2 Sep-Pak C18 Purification of Lysate Peptides

1. Add 1/20 volume of 20% TFA to the digest for a final concentration of 1% TFA. Check the pH by spotting a small amount of peptide sample on a pH strip (the pH should be under 3) (see Note 6). 2. After acidification, allow precipitate to form by letting sample stand on ice for 15 min. 3. Centrifuge the acidified peptide solution for 15 min at 1780
g at room temperature to remove the precipitate. 4. Transfer peptide-containing supernatant into a new tube without dislodging the precipitated material. 5. Purification of peptides is performed at room temperature on Sep-Pak C18 cartridges or columns from Waters Corporation (see Note 7). 6. Wet column with 500 μL Sep-Pak Wetting Solution (100% ACN). 7. Wash column with 1 mL of Sep-Pak Wash Solution A (0.1% TFA). 8. Repeat wash with 1 mL of Sep-Pak Wash Solution A (0.1% TFA). 9. Load peptides onto column by gravity flow (see Note 8). 10. Wash column with 1 mL of Sep-Pak Wash Solution A (0.1% TFA). 11. Repeat wash with 1 mL of Sep-Pak Wash Solution A (0.1% TFA). 12. Wash column with 500 μL Sep-Pak Wash Solution B (0.1% TFA/5% ACN). 13. Elute with 3
200 μL Sep-Pak Elution Solution (0.1% TFA/40% ACN) (see Note 9). 14. Dry samples in a speed vac with no heat, or freeze at 80  C overnight or in an ethanol/dry ice bath for 2 h and dry in a lyophilizer. 15. Once dry, samples can be stored at 80  C until ready to perform subsequent steps.

308

Anthony P. Possemato et al.

3.3 TMT Labeling of Peptides

1. Resuspend sample (i.e., 100 μg) in 70 μL 200 mM Na-HEPES, pH 8.5. 2. Equilibrate TMT reagents at room temperature for 15 min. 3. Add 30 μL of 100% ACN to TMT reagent (200 μg) and immediately add to sample. 4. Mix each sample with a brief and gentle vortex followed by brief centrifugation. 5. Incubate samples for 1 h at room temperature. Remove 1–2 μL of each sample for labeling efficiency check. 6. Labeling Efficiency Check: (a) Add 10
the sample volume with 0.1% TFA. (b) Check sample pH. If the pH is not 95% to continue protocol. 7. Quench the reaction by adding 50% hydroxylamine 1:10 (v:v) to a final concentration of 5%. 8. Vortex, spin down, and briefly incubate before placing in a vacuum concentrator without heat until dry.

3.4 C18 Sample Desalting [27] (See Note 10)

1. Pack a p10 micropipette tip with 2 Empore C18 circles cut from the disk using an 18 gauge blunt tip needle. Cut end off tip to remove excess plastic. Alternatively, use Pierce C18 Spin Tips. 2. Equilibrate the C18 Tip with 50 μL of C18 tip Wetting Solution (50% ACN, 0.1% TFA). 3. Wash C18 tip with 50 μL C18 tip Equilibrate & Wash Solution (0.1% TFA) 2
. Move tip to a new collection tube. 4. Load peptide sample on C18 tip, passing sample over tip 2
. 5. Wash C18 tip with 55 μL C18 tip Equilibrate & Wash Solution (0.1% TFA) 2
. Move tip to a new collection tube (see Note 11). 6. Elute peptides off the C18 tip with 10 μL of C18 tip Elution Solution (40% ACN, 0.1% TFA) 2
. Combine eluates into autosampler insert.

Total Proteome Mass Spectrometry Analysis

309

7. Dry down the C18 eluate in a speed vac and re-dissolve the peptides in an appropriate solvent for LC-MS analysis such as LCMS Solvent A (5% acetonitrile, 0.125% formic acid). 3.5 Basic Reversed Phase Fractionation

1. Resuspend elution in appropriate amount of bRP Solvent A (5% acetonitrile in 10 mM ammonium formate pH 10). Load 100–300 μg of combined sample on column. 2. Run sample over C18 column (i.e., Agilent Extend-C18, 5 μm, 2.1
150 mm) using a gradient from 5% to 45% bRP Solvent B with a flow rate of 200 μL/min. 3. Collect 96 fractions over the entire gradient and concatenate into 12 before drying down in a vacuum concentrator (see Note 12) (Fig. 1). 4. Desalt concatenated fractions over C18 tip (see Subheading 3.4) and redissolve the peptides in an appropriate solvent for LC-MS analysis such as LCMS Solvent A (3% acetonitrile, 0.12% formic acid).

3.6 LC-MS/MS Analysis (See Notes 13 and 14)

1. Resuspend peptides in 12–16 μL LCMS Solvent A (3% acetonitrile, 0.12% formic acid). 2. Load samples onto HPLC or UPLC system (Easy-nLC 1200 or equivalent) using a 50 cm
100 μM PicoFrit capillary column packed with C18 reversed-phase resin. 3. Elute peptides with a 150 min linear gradient of acetonitrile in 0.12% formic acid delivered at 280 nL/min (5–36% acetonitrile). 4. Collect data in the mass spectrometer using a CID-MS2/ HCD-MS3 method if available. 5. Instrument parameters: Method duration 210 min, user defined lock mass 371.10123, Orbitrap resolution 120 K, scan range 350–1400 m/z, maximum injection time 100 ms, AGC target 5E5, dynamic exclusion ¼ 1, exclusion duration 120 s, mass tolerance  7 ppm, MS2 isolation mode quadrupole, MS2 isolation window 0.4, activation type CID, collision energy mode fixed, collision energy 35, detector type IonTrap, max injection time 10 ms, AGC target 2E4, MS3 isolation mode quadrupole, isolation window 0.7, multi notch isolation, MS2 isolation window 3 m/z, number of notches ¼ 10, collision energy mode fixed, collision energy 65, detector type Orbitrap, Orbitrap resolution 50 K, max injection time 150 ms, AGC target 2.5E5.

3.7

Data Analysis

1. Process raw files, search, and score filter using preferred software package such as GFY-Core from Harvard University, which uses SEQUEST for database searching [9, 16, 28] (see Note 15).

310

Anthony P. Possemato et al.

2. Use the most recent update of the appropriate species database from Uniprot or NCBI (see Note 16). 3. Search using a target-decoy strategy with peptides in both forward and reverse directions, allowing for estimation of false discovery rate (FDR). 4. Database search settings: Mass accuracy  50 ppm for precursor ions. Mass accuracy  0.02 Da for product ions. Cysteine carboxamidomethylation is specified as a static modification (C + 57.02146374) as is TMT labeling of lysine (K + 229.162932) and TMT labeling of N termini (Nterm + 229.162932). Oxidation of methionine (M + 15.9949146221) is specified as a variable modification. 5. Filter results to a 1% peptide-level FDR. Further filter to a 1% protein-level FDR using a module such as ProteinSieve in GFY-Core. 6. Perform quantification by extracting TMT reporter ion signal: noise values for each identified peptide. Filter using minimum signal: noise of 189: 1. Depending on experimental design, a minimum number of TMT channels with TMT reporter ion signal may also be used as a filter (see Note 17). 7. Normalize total signal: noise across each channel by dividing the maximum signal: noise across the channels by the total signal: noise in each channel and multiplying each individual channel by the resulting number (so the offset is 1.0 for the max channel). 8. Sum the signal: noise values for each peptide representing a given protein. Generate fold changes between control sample and each treatment sample, or between each sample and an average across all channels.

4

Notes 1. Avoid foaming of samples in both homogenization (tissues) and sonication (cells or tissues) steps. 2. Centrifugation is performed at room temperature to avoid precipitation of urea. 3. Any protein assay used for normalizing total protein between samples should be compatible with the 9 M Urea Lysis Buffer. 4. DTT reduction should be performed at room temperature to avoid possible carbamylation of the proteins from breakdown of urea. 5. Digestion efficiency can be checked by reserving a small amount of predigested lysate and comparing to postdigestion lysate using a Bradford protein assay.

Total Proteome Mass Spectrometry Analysis

311

6. Before loading the peptides from the digested sample on the column, they must be acidified with TFA for efficient peptide binding. The acidification step helps remove fatty acids from the digested peptide mixture. 7. For larger sample sizes, higher capacity columns may be used (e.g., 360 mg SEP-PAK Classic C18, WAT051910) and the volume of wash and elution buffers should be adjusted accordingly. 8. Although some SEP PAK cartridges are reported to perform well under vacuum flow, optimal sample binding for all cartridges occurs with gravity flow only. Gentle vacuum pressure can be applied if samples do not flow under gravity. 9. Be sure that elution buffer is either prepared fresh immediately prior to elution or that it is stored in a vessel with a gasket that will prevent evaporation of the acetonitrile. 10. Tubes containing small volumes of acetonitrile-containing solutions should be prepared immediately before use and should be kept capped as much as possible, because the organic components evaporate quickly. 11. It is critical that the C18 tip is moved to a new tube at this step to collect the eluate separate from all previous loading/wash buffers. 12. Fractions are combined nonsequentially to maximize coverage of the proteome, making for more equal numbers of peptides in each fraction. See Fig. 1. 13. Preparation of peptides as described above provides sufficient material for ~10 injections on the instrument. 14. Run samples on a Thermo Orbitrap-Fusion Lumos Tribrid Mass spectrometer or equivalent with multinotch MS3 quantification enabled. 15. Many options exist for database searching and score filtering, including MaxQuant, Mascot, ProteomeDiscoverer, TransProteomicPipeline, etc. Individual preference for a particular software solution will determine which to use. 16. In some cases, one database will be more complete than another, and in rare cases there may be other resources for the most complete current database other than UniProt or NCBI. 17. Minimum number of channels with reporter ion signal is appropriate in cases where multiple channels would be expected to have signal, such as treatments of the same cell line, or experiments with biological replicates or triplicates. Experiments with a single channel for different cell lines or tissues may have very different repertoires of peptides and therefore this filter is not appropriate.

312

Anthony P. Possemato et al.

References 1. Bittar EE, Rivett AJ (1998) Intracellular protein degradation. In: Advances in molecular and cellular biology, vol 27, 1st edn. Jai Press Inc., Stamford, Connecticut 2. Hinkson IV, Elias JE (2011) The dynamic state of protein turnover: It’s about time. Trends Cell Biol 21(5):293–303 3. Popovic D, Vucic D, Dikic I (2014) Ubiquitination in disease pathogenesis and treatment. Nat Med 20(11):1242–1253 4. Vilchez D, Saez I, Dillin A (2014) The role of protein clearance mechanisms in organismal ageing and age-related diseases. Nat Commun 5:5659 5. Wang M, Kaufman RJ (2016) Protein misfolding in the endoplasmic reticulum as a conduit to human disease. Nature 529(7586):326–335 6. Gu S, Cui D, Chen X, Xiong X, Zhao Y (2018) PROTACs: an emerging targeting technique for protein degradation in drug discovery. BioEssays 40(4):1700247 7. Chamberlain PP, Hamann LG (2019) Development of targeted protein degradation therapeutics. Nat Chem Biol 15:937–944 8. Schapira M, Calabrese MF, Bullock AN, Crews CM (2019) Targeted protein degradation: expanding the toolbox. Nat Rev Drug Discov 18(12):949–963 9. Huttlin EL, Jedrychowski MP, Elias JE, Goswami T, Rad R, Beausoleil SA, Villen J, Haas W, Sowa ME, Gygi SP (2010) A tissuespecific atlas of mouse protein phosphorylation and expression. Cell 143:1174–1189 10. Mathieson T, Franken H, Kosinski J, Kurzawa N, Zinn N, Sweetman G, Poeckel D, Ratnu VS, Schramm M, Becher I, Steidel M, Noh KM, Bergamini G, Beck M, Bantscheff M, Savitski MM (2018) Systematic analysis of protein turnover in primary cells. Nat Commun 9 (1):689 11. Mertins P, Tang LC, Krug K, Clark DJ, Gritsenko MA, Chen L, Clauser KR, Clauss TR, Shah P, Gillette MA, Petyuk VA, Thomas SN, Mani DR, Mundt F, Moore RJ, Hu Y, Zhao R, Schnaubelt M, Keshishian H, Monroe ME, Zhang Z, Udeshi ND, Mani D, Davies SR, Townsend RR, Chan DW, Smith RD, Zhang H, Liu T, Carr SA (2018) Reproducible workflow for multiplexed deep-scale proteome and phosphoproteome analysis of tumor tissues by liquid chromatography-mass spectrometry. Nat Protoc 13(7):1632–1661 12. Savitski MM, Zinn N, Faelth-Savitski M, Poeckel D, Gade S, Becher I, Muelbaier M, Wagner AJ, Strohmer K, Werner T,

Melchert S, Petretich M, Rutkowska A, Vappiani J, Franken H, Steidel M, Sweetman GM, Gilan O, Lam EYN, Dawson MA, Prinjha RK, Grandi P, Bergamini G, Bantscheff M (2018) Multiplexed proteome dynamics profiling reveals mechanisms controlling protein homeostasis. Cell 173(1):260–274 13. Sperling AS, Burgess M, Keshishian H, Gasser JA, Bhatt S, Jan M, Słabicki M, Sellar RS, Fink EC, Miller PG, Liddicoat BJ, Sievers QL, Sharma R, Adams DN, Olesinski EA, Fulciniti M, Udeshi ND, Kuhn E, Letai A, Munshi NC, Carr SA, Ebert BL (2019) Patterns of substrate affinity, competition, and degradation kinetics underlie biological activity of thalidomide analogs. Blood 134 (2):160–170 14. Ong SE, Blagoev B, Kratchmarova I, Kristensen DB, Steen H, Pandey A, Mann M (2002) Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics 1(5):376–386 15. Thompson A, Sch€afer J, Kuhn K, Kienle S, Schwarz J, Schmidt G, Neumann T, Hamon C (2003) Tandem mass tags: a novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS. Anal Chem 75(8):1895–1904 16. Villen J, Beausoleil SA, Gerber SA, Gygi SP (2007) Large-scale phosphorylation analysis of mouse liver. Proc Natl Acad Sci U S A 104:1488–1493 17. Treumann A, Thiede B (2010) Isobaric protein and peptide quantification: perspectives and issues. Expert Rev Proteomics 7(5):647–653 18. Zhu W, Smith JW, Huang CM (2010) Mass spectrometry-based label-free quantitative proteomics. J Biomed Biotechnol 2010:840518 19. Tolonen AC, Haas W (2014) Quantitative proteomics using reductive dimethylation for stable isotope labeling. J Vis Exp 1:89 20. Stokes MP, Farnsworth CL, Gu H, Jia X, Worsfold CR, Yang V, Ren JM, Lee KA, Silva JC (2015) Complementary PTM profiling of drug response in human gastric carcinoma by Immunoaffinity and IMAC methods with Total proteome analysis. Proteomes 3(3):160–183 21. Hsu JL, Chen SH (2016) Stable isotope dimethyl labelling for quantitative proteomics and beyond. Philos Trans A Math Phys Eng Sci 374(2079):20150364 22. Rush J, Moritz A, Lee KA, Guo A, Goss VL, Spek EJ, Zhang H, Zha XM, Polakiewicz RD, Comb MJ (2005) Immunoaffinity profiling of

Total Proteome Mass Spectrometry Analysis tyrosine phosphorylation in cancer cells. Nat Biotechnol 23(1):94–101 23. Potu H, Peterson LF, Kandarpa M, Pal A, Sun H, Durham A, Harms PW, Hollenhorst PC, Eskiocak U, Talpaz M, Donato NJ (2017) Usp9x regulates Ets-1 ubiquitination and stability to control NRAS expression and tumorigenicity in melanoma. Nat Commun 8:14449 24. Ting L, Rad R, Gygi SP, Haas W (2011) MS3 eliminates ratio distortion in isobaric multiplexed quantitative proteomics. Nat Methods 8(11):937–940 25. McAlister GC, Nusinow DP, Jedrychowski MP, Wuhr M, Huttlin EL, Erickson BK, Rad R, Haas W, Gygi SP (2014) MultiNotch MS3 enables accurate, sensitive, and multiplexed detection of differential expression

313

across cancer cell line proteomes. Anal Chem 86(14):7150–7158 26. Paulo JA, O’Connell JD, Gygi SP, Triple A (2016) Knockout (TKO) proteomics standard for diagnosing ion interference in isobaric labeling experiments. J Am Soc Mass Spectrom 27(10):1620–1625 27. Rappsilber J, Ishihama Y, Mann M (2003) Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics. Anal Chem 75(3):663–670 28. Eng JK, McCormack AL, Yates JR (1994) An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J Am Soc Mass Spectrom 5:976–989

Chapter 17 PHOTACs Enable Optical Control of Protein Degradation Martin Reynders and Dirk Trauner Abstract Proteolysis Targeting Chimeras (PROTACs) are a promising technology to degrade specific target proteins. As bifunctional small molecules, PROTACs induce the ternary complex formation between an E3 ligase and a protein of interest (POI), leading to polyubiquitylation and subsequent proteasomal degradation of the protein in a catalytic fashion. We have developed a strategy to control PROTACs with the spatiotemporal precision of light, which led to light-activated versions, termed PHOTACs (PHOtochemically TArgeted Chimeras). By incorporating an azobenzene photoswitch into the PROTAC, we can reversibly control degradation of the POI, as demonstrated for BRD2-4 and FKBP12. Here, we describe our modular approach and the application of PHOTACs for the optical control of protein levels in detail. PHOTACs hold promise as both research tools and precision pharmaceutics. Key words PHOTAC, PROTAC, Photopharmacology, Chemical Optogenetics, Photocontrol, Light-activation, Photoswitch, Photochromic ligand

1

Introduction Cellular protein homeostasis is the tightly regulated steady state resulting from continuous synthesis and degradation of proteins. Small-molecule therapeutics targeting the cellular protein homeostasis machinery have been developed. These drugs mostly influence general functions, such as transcription, translation, or degradation, but do not target specific proteins. Proteolysis targeting chimeras (PROTACs) emerged recently as a new pharmacological strategy [1–4]. By connecting a ligand for an E3 ubiquitin ligase with a second one that targets a protein of interest (POI), these bifunctional molecules induce the formation of a productive ternary complex, which then leads to the polyubiquitylation at certain lysines of the POI. The targeted/polyubiquitylated POI is then recognized and degraded by the proteasome. The exact nature of the linker bridging both PROTAC ligands is crucial to ensure optimal complex assembly and protein degradation efficacy (“linkerology”).

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_17, © Springer Science+Business Media, LLC, part of Springer Nature 2021

315

316

Martin Reynders and Dirk Trauner

The first reported PROTACs recruited POIs to E3 ligases using peptides, [5–7] but recently synthetic ligands for E3 ligases, such as hydroxyproline ligands targeting the von Hippel–Lindau tumor suppressor (VHL) [8–10] and thalidomide-derivatives binding cereblon (CRBN), were developed [11, 12]. PROTACs have proven to be highly versatile and were found to be applicable to many protein classes [4]. Instead of merely inhibiting target activity like most classical drugs, PROTACs catalyze the degradation of proteins, thereby controlling their cellular concentrations. However, their fundamentally different mechanism of action (event-drivenvs. occupancy-driven) poses certain risks since the protein is removed with all of its functions. For example, BRD2 and BRD4 knockouts are lethal, whereas pharmacological inhibition is tolerated [13–15]. Consequently, it would be beneficial to activate PROTACs locally in cells and tissues where their effects (e.g., cytotoxicity) are desirable, while avoiding systemic toxicity. Photopharmacology is a method to spatiotemporally control drug activity with light and offers this additional layer of selectivity. Optical control can be achieved through different strategies: with caged compounds, [16, 17] engineered photoreceptors (Optogenetics), [18, 19] or by employing synthetic photoswitches, which can undergo reversible photochemical and thermal isomerization [20–22]. Here, we provide a protocol how to apply photopharmacology to targeted protein degradation. We have created PHOTACs through incorporation of an azobenzene photoswitch into the molecular structure of PROTACs (Fig. 1). Our lead PHOTACs show no pronounced effect on proteins levels in the dark but turn into active PROTACs when irradiated with violet-blue light (380–440 nm) (Fig. 3a, b). We have established light-dependent degradation and thus control of BRD2-4 and FKBP12 levels, by forming the ternary CRL4CRBN E3 ligase-POI complex and promoting polyubiquitylation upon irradiation. This optical control of protein levels translates to control of cell proliferation and viability in the case of BRD2-4 [23]. We aimed to design PHOTACs to be diversifiable and highly modular, while ensuring efficient synthetic access. To achieve this, we incorporated the photoswitch into the ligand for the E3 ligase and/or into the linker (Fig. 2a) [23]. This design does not restrict the POI ligand and allows us to control the formation of a productive ternary complex with light. We initially chose (+)-JQ1 as targeting ligand for BET proteins BRD2-4 (Fig. 3a–d) and the SLF ligand to target FKBP12 (Fig. 3e, f) [11]. PHOTAC-I-3 emerged as most effective lead compound from a small library of molecules targeting BRD2-4. PHOTAC-I-3 shows efficient photoswitching, as well as chemical stability. (Fig. 2a–e) The highest amount of active cis isomer is generated

PHOTACs Enable Optical Control of Protein Degradation

317

Fig. 1 Schematic depiction of PHOTACs and their catalytic mechanism of action to induce target protein degradation

at 390 nm (>90% cis), but similar photostationary states (PSS) can be obtained between 380 and 400 nm (Fig. 2c). Rapid deactivation via cis to trans isomerization can be performed by irradiation with wavelengths >450 nm, obtaining PSS of ca >70% trans (Fig. 2c). In the dark, cis PHOTAC-I-3 slowly isomerizes back to its inactive trans form with a half-life of 8.8 h at 37  C in DMSO (Fig. 2d). Validating our modular concept, PHOTAC-II-5 achieved a comparably strong degradation of FKBP12 upon 390 nm irradiation and was found to be inactive in the dark too (Fig. 3e, f). Since our initial disclosure of PHOTACs, irreversibly photoactivatable PROTACs have emerged [24, 25]. Masking PROTACs with a large photocleavable protecting group only allows for a

318

Martin Reynders and Dirk Trauner

A

B

dark 370 nm 380 nm 390 nm 400 nm 410 nm 420 nm 430 nm 440 nm 450 nm

0.4

Abs

0.3 0.2 0.1 0.0 300

400

500

600

460 nm 470 nm 480 nm 490 nm 500 nm 510 nm 520 nm 530 nm 540 nm 550 nm 700

Wavelength (nm)

C

E

D

0.6 0.4

0.2

PHOTAC-I-3 t1/2 (37 °C) = 8.76 ± 0.02 h

0.1

0.2 400

450

500

550

0.0

Abs (370 nm)

Abs (370 nm)

Fr action of trans

0.8

0.0

0.3

0.3

1.0

0.2

0.1

0.0 0

200

400

600

0

100

t (min)

Wavelength (nm)

200

300

t (min)

F

dark 370 nm 380 nm 390 nm 400 nm 410 nm 420 nm 430 nm 440 nm 450 nm

0.6

Abs

0.4

0.2

0.0 300

400

500

600

Wavelength (nm)

460 nm 470 nm 480 nm 490 nm 500 nm 510 nm 520 nm 530 nm 540 nm 550 nm 700

H 0.5 0.4

Abs (370 nm)

G

0.3 0.2 Photac-II-5 0.1 0.0

t1/2 (37 °C)= 6.88 ± 0.02 h 0

200

400

600

t (min)

Fig. 2 Photophysical properties, switching, and bistability of PHOTACs. (a) Switching of PHOTAC-I-3 between the trans isomer (left) and cis isomer (right). (b) UV-VIS spectra PHOTAC-I-3 following irradiation with different wavelengths for 5 min. (c) Fraction of trans PHOTAC-I-3 in the PSS. (d) Thermal relaxation of cis PHOTAC-I3 at 37  C in DMSO. (e) Reversible switching and photochemical stability of PHOTAC-I-3. (f) Switching of PHOTAC-II-5 between the trans isomer (left) and cis isomer (right). (g) UV-VIS spectra PHOTAC-II-5 following irradiation with different wavelengths for 5 min. (h) Thermal relaxation of cis PHOTAC-II-5 at 37  C in DMSO. (Reproduced from Reynders et al., 2020. Copyright 2020. The Authors, some rights reserved; exclusive licensee AAAS. Distributed under a Creative Commons Attribution License 4.0 (CC BY) [23])

single controlled activation event, after which the free PROTAC is obtained and needs to be inactivated through diffusion or metabolism. A major advantage of photoswitches over classical compounds is their reversibility by thermal or photochemical isomerization to

PHOTACs Enable Optical Control of Protein Degradation

319

their inactive state, which allows control of the activity of photoswitches at all times. The thermal and photochemical reversibility can be demonstrated in a rescue experiment, wherein PHOTAC-I3 was activated for 1 min (390 nm irradiation) and then either left in the dark or pulse irradiated (100 ms every 10 s) with deactivating green light (525 nm). Initial BRD2 degradation is observed in RS4;11 cells, but protein levels can recover, and do so faster when irradiated with green light (Fig. 3d). Optogenetic approaches to protein degradation require engineering and expression of a photosensitive LOV2 domain-fused degron [26–29]. Although it works well in vivo, this method to study biological pathways perturbs the protein activity and may create nonphysiological protein distribution and levels. PHOTACs, by contrast, operate like drugs in native tissues. The present chapter provides a detailed procedure to employ CRBN-based PHOTACs for the light control of either BET proteins (BRD2-4) or FKBP12 levels. This modular approach will guide the future development of PHOTACs targeting other protein classes. A second topic is the development of red-shifted and fast relaxing photoswitches to boost the temporal and spatial precision of dark-inactive compounds [30–33]. We and others [34] anticipate that PHOTACs will become important tools in cell biology, and we also envision clinical applications. A major risk for toxicity of PROTACs is their catalytic mechanism of action. PHOTACs could circumvent this risk through spatiotemporally confined activation within a specific area or tissue.

2

Materials

2.1

Cell Culture

A widely used cell line for PROTACs and PHOTACs is the human acute lymphoblastic leukemia RS4;11 cell line (ATCC® CRL1873TM). Cells are maintained in phenol red free RPMI1640 medium (Gibco 11835030) with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (PS) in a humidified incubator at 37  C with 5% CO2 in air. Cell count is maintained between 105 and 106 cells per milliliter.

2.2

Reagents

PHOTACs are stored in the dark and prepared as previously described from unpurified commercial reagents. DMSO, MeOH, and PBS (pH 7.4) are used as purchased.

2.3

Light Source

Pulsed irradiation of cells over long periods of time can be performed with a cell disco [35], a plate of 5 mm LEDs controlled by an Arduino UNO or similar cheap computer. (Fig. 4). For the experiments described here, an Arduino UNO is connected to a 5 V power source and used to control a 5 V relay

320

Martin Reynders and Dirk Trauner

A

B dark

590 nm

572 nm

545 nm

525 nm

505 nm

490 nm

465 nm

430 nm

450 nm

BRD4

PHOTAC-I-3 (300 nM)

DMSO 10 μM 3 μM 1 μM 300 nM 100 nM 30 nM 10 nM 1 μM + MLN

DMSO 10 μM 3 μM 1 μM 300 nM 100 nM 30 nM 10 nM 1 μM + MLN

410 nm

PHOTAC-I-3

PHOTAC-I-3

390 nm

Dark 370 nm

390 nm

BRD4

BRD2

γTUBULIN

BRD3

BRD4

DMSO

BRD2 c-MYC

BRD2

γTUBULIN

CUL4A

γTUBULIN

390 nm - 1min, then Dark

390 nm - 1min, then 525 nm

24 h

24 h

12 h

6h

3h

1.5 h

PHOTAC-I-3 - 300 nM DMSO

0h

24 h

DMSO

24 h

12 h

6h

3h

1.5 h

PHOTAC-I-3 - 300 nM

0h

12 h

DMSO 24 h

3h

0h

24 h

1.5 h

PHOTAC-I-3 - 1 μM

DMSO 24 h

12 h

6h

3h

1.5 h

PHOTAC-I-3 - 1 μM 0h

D

Dark

24 h

390 nm

6h

C

BRD4

BRD2

BRD3

PCNA

BRD2 c-MYC PARP1 cleaved PARP1 PCNA

Dark

390 nm

1 μM + MLN

10 nM

30 nM

100 nM

3 μM

10 μM

DMSO

1 μM + MLN

10 nM

30 nM

100 nM

300 nM

1 μM

3 μM

10 μM

DMSO

1 μM

PHOTAC-II-5

PHOTAC-II-5

300 nM

E

FKBP12 CUL4A γTUBULIN

Dark DMSO

24 h

12 h

3h

1.5 h

0h

24 h

24 h

12 h

6h

3h

1.5 h

0h

PHOTAC-II-5 - 100 nM

24 h

390 nm PHOTAC-II-5 - 100 nM DMSO

6h

F

FKBP12 γTUBULIN

Fig. 3 Optical control of BRD2–4 and FKBP12 levels. (a) Immunoblot analysis after treatment of RS4;11 cells with PHOTAC-I-3 for 4 h at different concentrations. Cells were either irradiated with 100 ms pulses of 390 nm light every 10 s (left) or kept in the dark (right). (b) Time course of BRD2–4 degradation, c-MYC levels, and

PHOTACs Enable Optical Control of Protein Degradation

321

module. The relay connects the 24 V (2 A) power supply with the LEDs to switch them on and off (Fig. 4a). A previously created program [35] can be used to control the Arduino, and a pulsed irradiation of 100 ms every 10 s is chosen (1% duty cycle) to provide constant irradiation, while keeping overall light exposure low and avoiding excessive heating. 5 mm LED boards with 24 LEDs were used for the pulsed irradiation of 6 well plates (Fig. 4b) [35]. These consist of 4 parallel rows each made with slots for 6 LEDs each and a resistor (47 Ω, 1 W). The LEDs are aligned to a 24-well plate to evenly distribute them across the plate. Rubber feet ensure proper spacing between the LEDs and the well plate. LED plates are connected to the Arduino UNO using standard copper wires and can be placed in a humidified incubator. 5 mm LEDs were obtained from Roithner Lastertechnik (370 nm (XSL-370-5E), 390 nm (VL390-5-15), 410 nm (VL410-5-15), 430 nm (VL430-5-15), 450 nm (ELD-450-525), 465 nm (RLS-B465), 477 nm (RLS-5B475-5), 490 nm (LED49003), 505 nm (B5-433-B505), 525 nm (B5-433-B525), 545 nm (LED545-04), 572 nm (B5-433-20), and 590 nm (CY5111AWY)). Generally, LEDs with a small bandwidth are recommended to obtain the best isomer ratios, and a wide cone angle will ensure equal light distribution across the well plate. Recommended power output for the LEDs is 10–20 mW. Other light sources can be used to irradiate PHOTACs such as lasers, monochromators, or high-power LEDs. For the use of broad, polychromatic light sources, a narrow band-pass filter set is recommended for best results. Furthermore, a strong red LED light source with λ > 650 nm is recommended to prepare samples under “Dark”/red light conditions. Black, nonairtight plastic boxes (20  15  15 cm) are used to shield the controls from light.

ä Fig. 3 (continued) PARP1 cleavage assayed by immunoblotting. RS4;11 cells were treated with PHOTAC-I-3 (1μM) and collected at the indicated time points. PHOTAC-I-3 has no effect on BRD2–4 levels in the dark over several hours. (c) Immunoblot analysis of a rescue experiment demonstrating the reversibility of degradation promoted by PHOTAC-I-3 through thermal relaxation (left) or optical inactivation by 525 nm pulsed irradiation (right, 100 ms every 10 s). (d) Color dosing: Wavelength dependence of BRD2/4 degradation promoted by 300 nM PHOTAC-I-3. (e) Immunoblot analysis of FKBP12 after treatment of RS4;11 cells with PHOTAC-II-5 for 4 h at different concentrations. Cells were either irradiated with pulses of 390 nm light (left, 100 ms every 10 s) or kept in the dark (right). (f) Time course of FKBP12 degradation visualized by immunoblotting. RS4;11 cells were treated with PHOTAC-II-5 (100 nM) and collected at the indicated time points. (Reproduced from Reynders et al., Copyright 2020 The Authors, some rights reserved; exclusive licensee AAAS. Distributed under a Creative Commons Attribution License 4.0 (CC BY) [23])

322

Martin Reynders and Dirk Trauner

Fig. 4 Cell disco. (a) Programmable Arduino Uno -1- and relay switch -2- setup to control the light source. (b) 24-LED plate with 6-well plate used for cell-culture irradiation 2.4

Immunoblotting

1. Lysis buffer is prepared fresh by mixing 1.948 mL of RIPA lysis buffer (stored at 4  C) with 40μL of a 50 protease inhibitor cocktail stock (cOmplete™ Protease Inhibitor Cocktail, Roche, 1 tablet dissolved in 1 mL of RIPA, stored at 20  C) and 12μL of phosphatase inhibitor cocktail (Millipore Sigma). 2. 4 Laemmli sample buffer can be prepared by adding 100μL of 2-mercaptoethanol per 900μL of 4 Laemmli sample buffer (BIO-RAD) and is stored at 20  C.

PHOTACs Enable Optical Control of Protein Degradation

323

3. SDS Page running buffer is prepared by mixing 50 mL of Invitrogen™ Novex™ NuPAGE™ MES SDS Running Buffer (20) and 950 mL of purified water. 4. Membrane transfer buffer (TAE buffer containing 20%v/v MeOH) is prepared by adding 1.4 L of purified water to 200 mL of a 10 TAE buffer stock, followed by 400 mL of MeOH. 5. Washing buffer (TBST) is prepared by dissolving tris base (24 g) and NaCl (175 g) in 20 L of purified water, adjusting the pH to 7.4 using HCl and adding 20 mL of Tween 20. After stirring for 30 min, the buffer can be used. 6. Blocking buffer is prepared by suspending 25 g of blotting grad milk powder in 500 mL of washing buffer (TBST) and can be stored at 4  C. 7. Protein concentration determination assay kit such as Pierce™ BCA Protein Assay Kit. 8. HRP substrate such as SuperSignal™ West Pico PLUS Chemiluminescent Substrate. 9. PVDF membrane. 10. Precast SDS page gels (such as NuPAGE™ 4–12% Bis-Tris Midi Protein Gels, 26-well).

3

Methods

3.1 Preparation of PHOTAC Stock Solutions

3.2 PHOTAC Treatment

PHOTACs should be stored in the dark and are dissolved in DMSO under red LED light (λ > 650 nm) conditions to create 10 mM stock solutions. The stock solutions should always be stored in the dark to avoid isomerization by ambient light. Storage in the dark will also promote the thermal relaxation of the active cis- to the inactive trans-PHOTACs. 1. Prior to treatment, place the black plastic boxes and LED plates in the incubator. Add 2  106 RS4;11 cells in 2 mL of phenol red free RPMI1640 medium to each of the 6-well plates and allow to equilibrate in the incubator. 2. Dilute PHOTACs to prepare 2 stocks. Perform these steps 2 and 3 under red or low light conditions for best results. For example, dissolve 10μL of PHOTAC-I-3-stock (10 mM) in 90μL of DMSO, add 4.9 mL of warm phenol red free RPMI1640 medium, and mix thoroughly. Tenfold serial dilutions are prepared by mixing 500μL of the previous dilation with 4.5 mL of phenol red free RPMI1640 medium containing 2% DMSO. For most PHOTACs, we recommend an ideal working concentration between 100 nM and 1μM, whereas

324

Martin Reynders and Dirk Trauner

the hook effect can be observed at concentrations 10μM (see Notes 1 and 3). 3. Add 2 mL of the diluted PHOTACs to the wells containing RS4;11 cells and place in the previously prepare boxes. Optionally, cells can be irradiated with the chosen wavelength, i.e., 390 nm, for 1 min to ensure a rapid isomerization right at the start. Cells are then irradiated using the cell disco to ensure a constant isomer ratio throughout the experiment (100 ms pulses every 10 s). For short experiments, a single strong irradiation event can be sufficient to active most PHOTACs, given their thermal stability. Strong degradation of target proteins BRD2-4 for PHOTAC-I-3 or FKBP12 for PHOTAC-II5 is observed within 4 h (see Notes 2–4). 4. After incubation, collect cells under red light. Carefully take up the medium using serological pipettes and wash the wells two times to resuspend all cells and centrifuge (200  g, 5 min) at 4  C Aspirate the supernatant, wash the pellets with ice cold PBS (1 mL), and spin down (400  g, 5 min) at 4  C. Optionally, cell pellets can be frozen for storage at 80  C at this point. 5. Lyse cells on ice for 20 min by the addition of 37 mL of RIPA lysis buffer containing protease and phosphatase inhibitor cocktails and suspend the cell pellet by repeated pipetting. 6. Spin down lysates at 15,000  g for 10 min at 4  C. 7. Carefully collect the supernatant using thin pipet tips for gel loading to avoid taking up the precipitate. 8. Determine the protein concentration in a 96-well plate, in triplicate (2μL filtrate each), using the BCA method, i.e., using a commercial kit such as the Pierce™ BCA Protein Assay Kit. 9. Add 10μL of 4 Laemmli buffer containing DTT and heat the samples at 95  C for 10 min. Optionally, samples can be sonicated at this point to avoid DNA contamination. 10. Put a precast gel into the gel cassette, add MES-SDS running buffer, and wash all pockets with running buffer. 11. Load 25μg of protein on a precast gel (Tris-glycine, Invitrogen) and run for 10 min at 100 V, followed by 70–120 min at 130 V (longer runtimes allow for better separation of high molecular weight bands, but low molecular weight proteins will be lost). 12. Prepare transfer buffer and fill the transfer chamber, activate a PVDF membrane in methanol for 5 min, and assemble the transfer stack.

PHOTACs Enable Optical Control of Protein Degradation

325

13. Perform membrane transfer at 30 V for 120 min. Optionally, membrane transfer can be performed in a cold room. 14. Wash membranes once briefly in purified water, stain using Ponceau S while shaking for 1 min, wash two to three times with purified water to remove excess Ponceau S, and place the membrane between 2 layers of plastic wrap. Cut membranes into small strip and block in blocking buffer for 30 min. 15. Wash membranes three times with washing buffer and incubate with primary antibodies overnight at 4  C on a shaker, diluted as per the supplier’s instructions in TBST with 3%BSA and 0.05% NaN3. 16. Subsequently, wash membranes twice quickly with washing buffer, followed by washing twice for 10 min to remove primary antibodies. 17. Add secondary HRP-conjugated antibodies, diluted in blocking buffer (1:5000), and incubate for 30 min at RT. 18. Wash membranes twice quickly with washing buffer, followed by washing twice for 10 min, and then transfer to PBS until imaged. 19. Remove membrane strips from PBS, carefully remove excess PBS on a piece of paper towel, place on a flat surface, and add the premixed HRP substrate solution. 20. Image chemoluminescence in incremental steps, using a GE ImageQuant LAS 4000 series or equivalent imaging system.

4

Notes 1. Working with photochromic compounds: It is important to consider their photophysical behavior at all times. Azobenzenes effectively absorb light in the visible region of the electromagnetic spectrum and hence isomerize upon exposure to ambient light conditions. The isomer ratio is strongly dependent on the light source as well as the particular spectral overlap and quantum efficiencies of isomerization of the E- and Zisomers. For the PHOTACs presented here, irradiation with wavelengths above 600 nm is possible without influencing the isomer ratio. Hence, working under red light or orthogonal imaging using red fluorophores (excitation above 600 nm) is recommended for best results. Keep in mind that some light sources allow for second order diffraction and thus do not emit monochromatic light. Furthermore, we recommend recording emission spectra or obtaining detailed technical information on the light sources from the supplier to ensure reproducibility. PHOTACs thermally relax back to their inactive isomer, and accidental exposure could be reverted by storage in the

326

Martin Reynders and Dirk Trauner

dark. Note that this works best at elevated temperatures (i.e., 37  C). Activation of PHOTACs under saturating irradiation conditions will establish an equilibrium state between the isomers depending on the spectrum of the used light source. While 390 nm is the ideal wavelength for E-to-Z isomerization for the PHOTACs discussed herein, wavelength between 370 and 430 nm can activate them well and commercial “black lights” or common 405 nm excitation on microscopes can be used to activate PHOTACs. For most experiments, saturating irradiation conditions (100 ms pulses every 10 s) have been used. Less irradiation can be sufficient to activate PHOTACs, and the irradiation protocol can be adjusted freely. It is possible to work without a cell disco, just manually irradiate PHOTACs, for example, for approx. 30 s every hour, and still achieve good results, thanks to their thermal bistability. While the LED boards usually work without defects over many experiments, it is recommended to check the LEDs for proper function before and after the experiments, as one malfunctioning LED will disrupt the whole row. Given that phenol red is an azobenzene itself, it strongly absorbs in the visible spectrum and can even act as a photosensitizer. We recommend using phenol red free media at all times. 2. Choice of cell line: Cereblon expression is crucial for the function of the PHOTACs discussed here and knockdown/knockout of CRBN can be used as a negative control. However, the efficacy of PROTACs and PHOTAC can vary across different cell lines. While RS4;11 cells show fast degradation using PHOTACs, adherent cell lines such as MDA-MB-231, MDA-MB-468, or HEK293T show slower response to PHOTACs and require prolonged incubation times. If prolonged irradiation is not desirable, preincubation with PHOTACs is possible to reduce irradiation times. 3. Dosing and inactivation of PHOTACs: The commonly observed saturation of POI and the E3 ligase by PROTACs, termed hook-effect, is also observed for PHOTACs. Therefore, high concentration of PHOTACs will inhibit protein degradation, and we recommend using working concentrations below 3μM. We recommend using 2μM to 100 nM PHOTAC, although ideal concentrations can differ due to target protein expression, resynthesis rates, and cell permeability. Furthermore, it must be considered that PHOTACs in their inactive form can still act as inhibitors of their target proteins, while being unable to form a productive ternary complex that leads to protein ubiquitylation. This means that the inactive

PHOTACs Enable Optical Control of Protein Degradation

327

PHOTAC can act as a competitive binder, displacing the active PHOTAC. After activating PHOTACs with light, such as 390 nm irradiation, inactivation can either occur thermally, which is a slow process on the scale of hours for the current generation of PHOTACs or by irradiation with a second wavelength in the range of 500 to 560 nm (green light). Irradiation with green light induces the Z-to-E isomerization and shifts the equilibrium toward the inactive form. This Z-to-E isomerization is not complete, and a small fraction of active PHOTAC remains. The wavelength-dependent E/Z ratio of azobenzenes at saturating irradiation conditions, known as the photostationary state (PSS), can also be used to dose the amount of active PHOTAC and thus the amount of target protein, by using different colors of light. Starting with green light (e. g. 550 nm), only small amounts of PHOTAC are activated, but by moving toward blue to violet light (460 to 390 nm), larger fractions of PHOTAC are activated and, consequently, more protein is degraded. 4. Time course and other experiments: Such experiments are useful to observe changes in protein levels over time. To observe the light-induced protein degradation by PHOTACs over time, we recommend preparing an excess of 2 stock of PHOTAC in phenol red free medium for all timepoints at the start of the assay. Keep the stock in the dark and add it to the cell containing well plate at your chosen timepoints with a joint end point of the assay. Thereby, cells are irradiated for the same duration and samples are worked up under the same conditions. The cell disco can further be used to create more complex irradiation patterns and consequently protein level patterns. One can also place a second LED plate on top of the well plate and connect it to a second program [35] to modulate PHOTAC activity with different wavelengths. When irradiating from the top, refrain from labeling the well plate as a layer of black marker can absorb a significant amount of light. For color dosing, we recommend starting at long wavelength (600 nm) and slowly moving toward shorter wavelengths. References 1. Burslem GM, Crews CM (2017) Smallmolecule modulation of protein homeostasis. Chem Rev 117:11269–11301. https://doi. org/10.1021/acs.chemrev.7b00077 2. Lai AC, Crews CM (2017) Induced protein degradation: an emerging drug discovery

paradigm. Nat Rev Drug Discov 16:101–114. https://doi.org/10.1038/nrd.2016.211 3. Skaar JR, Pagan JK, Pagano M (2014) SCF ubiquitin ligase-targeted therapies. Nat Rev Drug Discov 13:889–903. https://doi.org/ 10.1038/nrd4432

328

Martin Reynders and Dirk Trauner

4. Sun X, Gao H, Yang Y, He M, Wu Y, Song Y, Tong Y, Rao Y (2019) PROTACs: great opportunities for academia and industry. Sig Transduct Target Ther 4:1–33. https://doi.org/10. 1038/s41392-019-0101-6 5. Sakamoto KM, Kim KB, Kumagai A, Mercurio F, Crews CM, Deshaies RJ (2001) Protacs: chimeric molecules that target proteins to the Skp1–Cullin–F box complex for ubiquitination and degradation. PNAS 98:8554–8559. https://doi.org/10.1073/ pnas.141230798 6. Sakamoto KM, Kim KB, Verma R, Ransick A, Stein B, Crews CM, Deshaies RJ (2003) Development of Protacs to target cancer-promoting proteins for ubiquitination and degradation. Mol Cell Proteomics 2:1350–1358. https:// doi.org/10.1074/mcp.T300009-MCP200 7. Schneekloth John S, Fonseca FN, Koldobskiy M, Mandal A, Deshaies R, Sakamoto K, Crews CM (2004) Chemical genetic control of protein levels: selective in vivo targeted degradation. J Am Chem Soc 126:3748–3754. https://doi.org/10.1021/ ja039025z 8. Bondeson DP, Mares A, Smith IED, Ko E, Campos S, Miah AH, Mulholland KE, Routly N, Buckley DL, Gustafson JL, Zinn N, Grandi P, Shimamura S, Bergamini G, Faelth-Savitski M, Bantscheff M, Cox C, Gordon DA, Willard RR, Flanagan JJ, Casillas LN, Votta BJ, den Besten W, Famm K, Kruidenier L, Carter PS, Harling JD, Churcher I, Crews CM (2015) Catalytic in vivo protein knockdown by smallmolecule PROTACs. Nat Chem Biol 11:611–617. https://doi.org/10.1038/ nchembio.1858 9. Zengerle M, Chan K-H, Ciulli A (2015) Selective small molecule induced degradation of the BET Bromodomain protein BRD4. ACS Chem Biol 10:1770–1777. https://doi.org/ 10.1021/acschembio.5b00216 10. Buckley DL, Raina K, Darricarrere N, Hines J, Gustafson JL, Smith IE, Miah AH, Harling JD, Crews CM (2015) HaloPROTACS: use of small molecule PROTACs to induce degradation of HaloTag fusion proteins. ACS Chem Biol 10:1831–1837. https://doi.org/10. 1021/acschembio.5b00442 11. Winter GE, Buckley DL, Paulk J, Roberts JM, Souza A, Dhe-Paganon S, Bradner JE (2015) Phthalimide conjugation as a strategy for in vivo target protein degradation. Science 348:1376–1381. https://doi.org/10.1126/ science.aab1433 12. Lu J, Qian Y, Altieri M, Dong H, Wang J, Raina K, Hines J, Winkler JD, Crew AP,

Coleman K, Crews CM (2015) Hijacking the E3 ubiquitin ligase Cereblon to efficiently target BRD4. Chem Biol 22:755–763. https:// doi.org/10.1016/j.chembiol.2015.05.009 13. Houzelstein D, Bullock SL, Lynch DE, Grigorieva EF, Wilson VA, Beddington RSP (2002) Growth and early Postimplantation defects in mice deficient for the Bromodomain-containing protein Brd4. Mol Cell Biol 22:3794–3802. https://doi.org/10. 1128/MCB.22.11.3794-3802.2002 14. Shang E, Wang X, Wen D, Greenberg DA, Wolgemuth DJ (2009) Double bromodomain-containing gene Brd2 is essential for embryonic development in mouse. Dev Dyn 238:908–917. https://doi.org/10. 1002/dvdy.21911 15. Gyuris A, Donovan DJ, Seymour KA, Lovasco LA, Smilowitz NR, Halperin ALP, Klysik JE, Freiman RN (2009) The chromatin-targeting protein Brd2 is required for neural tube closure and embryogenesis. Biochim Biophys Acta 1789:413–421. https://doi.org/10.1016/j. bbagrm.2009.03.005 16. Hansen MJ, Velema WA, Lerch MM, Szymanski W, Feringa BL (2015) Wavelengthselective cleavage of photoprotecting groups: strategies and applications in dynamic systems. Chem Soc Rev 44:3358–3377. https://doi. org/10.1039/C5CS00118H 17. Silva JM, Silva E, Reis RL (2019) Lighttriggered release of photocaged therapeutics where are we now? J Control Release 298:154–176. https://doi.org/10.1016/j. jconrel.2019.02.006 18. Fenno L, Yizhar O, Deisseroth K (2011) The development and application of Optogenetics. Annu Rev Neurosci 34:389–412. https://doi. org/10.1146/annurev-neuro-061010113817 19. Repina NA, Rosenbloom A, Mukherjee A, Schaffer DV, Kane RS (2017) At light speed: advances in optogenetic systems for regulating cell signaling and behavior. Ann Rev Chem Biomol Eng 8:13–39. https://doi.org/10. 1146/annurev-chembioeng-060816-101254 20. Beharry AA, Woolley GA (2011) Azobenzene photoswitches for biomolecules. Chem Soc Rev 40:4422–4437. https://doi.org/10. 1039/C1CS15023E 21. Lerch MM, Hansen MJ, van Dam GM, Szymanski W, Feringa BL (2016) Emerging targets in photopharmacology. Angew Chem Int Ed 55:10978–10999. https://doi.org/10. 1002/anie.201601931 22. Hu¨ll K, Morstein J, Trauner D (2018) In Vivo Photopharmacology. Chem Rev

PHOTACs Enable Optical Control of Protein Degradation 118:10710–10747. https://doi.org/10. 1021/acs.chemrev.8b00037 23. Reynders M, Matsuura BS, Be´routi M, Simoneschi D, Marzio A, Pagano M, Trauner D (2020) PHOTACs enable optical control of protein degradation. Sci Adv 6:eaay5064. https://doi.org/10.1126/sciadv.aay5064 24. Naro Y, Darrah K, Deiters A (2019) Optical control of small molecule-induced protein degradation. J Am Chem Soc 142(5):2193–2197. https://doi.org/10.26434/chemrxiv. 8216714.v1 25. Xue G, Wang K, Zhou D, Zhong H, Pan Z (2019) Light-induced protein degradation with Photocaged PROTACs. J Am Chem Soc 141(46):18370–18374. https://doi.org/10. 1021/jacs.9b06422 26. Renicke C, Schuster D, Usherenko S, Essen L-O, Taxis C (2013) A LOV2 domain-based Optogenetic tool to control protein degradation and cellular function. Chem Biol 20:619–626. https://doi.org/10.1016/j. chembiol.2013.03.005 27. Bonger KM, Rakhit R, Payumo AY, Chen JK, Wandless TJ (2014) General method for regulating protein stability with light. ACS Chem Biol 9:111–115. https://doi.org/10.1021/ cb400755b 28. Usherenko S, Stibbe H, Musco´ M, Essen L-O, Kostina EA, Taxis C (2014) Photo-sensitive degron variants for tuning protein stability by light. BMC Syst Biol 8. https://doi.org/10. 1186/s12918-014-0128-9 29. Hermann A, Liewald JF, Gottschalk A (2015) A photosensitive degron enables acute lightinduced protein degradation in the nervous system. Curr Biol 25:R749–R750. https:// doi.org/10.1016/j.cub.2015.07.040

329

30. Ble´ger D, Schwarz J, Brouwer AM, Hecht S (2012) O-Fluoroazobenzenes as readily synthesized Photoswitches offering nearly quantitative two-way isomerization with visible light. J Am Chem Soc 134:20597–20600. https://doi.org/10.1021/ja310323y 31. Dong M, Babalhavaeji A, Samanta S, Beharry AA, Woolley GA (2015) Red-shifting Azobenzene Photoswitches for in vivo use. Acc Chem Res 48:2662–2670. https://doi.org/10. 1021/acs.accounts.5b00270 32. Hansen MJ, Lerch MM, Szymanski W, Feringa BL (2016) Direct and versatile synthesis of red-shifted Azobenzenes. Angew Chem Int Ed 55:13514–13518. https://doi.org/10. 1002/anie.201607529 33. Dong M, Babalhavaeji A, Collins CV, Jarrah K, Sadovski O, Dai Q, Woolley GA (2017) Nearinfrared Photoswitching of Azobenzenes under physiological conditions. J Am Chem Soc 139:13483–13486. https://doi.org/10. 1021/jacs.7b06471 34. Pfaff P, Samarasinghe KTG, Crews CM, Carreira EM (2019) Reversible spatiotemporal control of induced protein degradation by Bistable PhotoPROTACs. ACS Cent Sci 5 (10):1682–1690. https://doi.org/10.1021/ acscentsci.9b00713 35. Borowiak M, Nahaboo W, Reynders M, Nekolla K, Jalinot P, Hasserodt J, Rehberg M, Delattre M, Zahler S, Vollmar A, Trauner D, Thorn-Seshold O (2015) Photoswitchable inhibitors of microtubule dynamics optically control mitosis and cell death. Cell 162:403–411. https://doi.org/10.1016/j. cell.2015.06.049

Chapter 18 Protocols for Synthesis of SNIPERs and the Methods to Evaluate the Anticancer Effects Yoshinori Tsukumo, Genichiro Tsuji, Hidetomo Yokoo, Norihito Shibata, Nobumichi Ohoka, Yosuke Demizu, and Mikihiko Naito Abstract Inducing degradation of undruggable target proteins by the use of chimeric small molecules, represented by proteolysis-targeting chimeras, is a promising strategy for drug development. We developed a series of chimeric molecules, termed “specific and nongenetic inhibitor of apoptosis protein (IAP)-dependent protein erasers” (SNIPERs) that recruit IAP ubiquitin ligases to induce degradation of target proteins. SNIPERs also induce degradation of some IAPs, including cIAP1 and XIAP, which are antiapoptotic proteins that are overexpressed in many cancers. Such protein degraders have unique properties that could be especially useful in cancer therapy. This chapter describes (1) the design and synthesis of SNIPER compounds, (2) the methods used for the detection of target protein degradation and ubiquitylation, and (3) the protocol to evaluate the antitumor activity of SNIPER. Key words SNIPER, Cancer, Degradation

1

Introduction Inducing degradation of specific proteins by chimeric small molecules is an emerging approach for drug development. We developed a protein knockdown system using a series of hybrid small molecules called “SNIPERs” (Specific and Nongenetic inhibitor of apoptosis protein [IAP]-dependent Protein ERasers), which induce target protein degradation via the ubiquitin-proteasome system (UPS) [1–6]. SNIPERs are bispecific compounds composed of two distinct ligands. One ligand interacts with E3 ubiquitin ligase of the IAP family, and the other is a ligand for the target protein. These two ligands are linked by spacers with optimal lengths. Accordingly, SNIPERs can make the target protein spatially available to the IAP ubiquitin ligase, thereby inducing IAP-mediated

Yoshinori Tsukumo and Genichiro Tsuji contributed equally to this work. Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9_18, © Springer Science+Business Media, LLC, part of Springer Nature 2021

331

332

Yoshinori Tsukumo et al.

ubiquitylation of the target protein and subsequent proteasomedependent degradation (Fig. 1). Importantly, several SNIPERs including SNIPER(ER), which induces degradation of estrogen receptor (ER), exhibit antitumor activity in vitro and in vivo [6–9, 14]. In this chapter, we describe the experimental methods used in these studies, focusing on (1) design and synthesis of SNIPERs; (2) evaluating the degradation of their target proteins; and (3) measuring the antitumor activity of SNIPER(ER)-87, as a representative compound. We designed and synthesized SNIPER(ER)-14, -19 and -87. These three compounds were developed by ligating 4-hydroxytamoxifen (an ERα antagonist used to treat ERα-positive breast cancer) with different IAP-binding compounds: bestatin, MV1, and LCL161 (Fig. 2). The design of these SNIPER(ER)s was based on the X-ray cocrystal structure (PDB: 3ERT) between 4-hydroxytamoxifen and ERα [10]. Each IAP ligand was bound to the dimethyl-amino moiety of 4-hydroxytamoxifen via an alkyl spacer, because this moiety is exposed on the ERα surface [8, 10]. Bestatin, an aminopeptidase inhibitor isolated from actinomycetes [11, 12], is a first-generation IAP ligand, whose IAP-binding ability was reported by us [13]. However, bestatinbased SNIPER(ER)-14 could only induce ERα degradation at 10 μM or higher [7]. We then developed SNIPER(ER)-19, in which the IAP antagonist MV1 was substituted for bestatin; it successfully improved ERα degradation activity [8]. We further developed SNIPER(ER)-87, with higher degradation activity, by using a derivative of LCL161 [8], which is an IAP antagonist under phase I study. Similarly, the improved degradation activity was also observed with SNIPER(ABL)s, which conjugated BCR-ABL inhibitors with IAP antagonists including LCL161 [14]. A schematic to synthesize these SNIPERs is shown in Figs. 2 and 3. We also developed SNIPER(BRD4)-1 and SNIPER(PDE4)-9 against different therapeutic target proteins, BRD4, and PDE4, by conjugating the LCL161 derivative to a BRD4 inhibitor (JQ-1) and to a PDE4 inhibitor, respectively [8]. A schematic for synthesizing these SNIPERs is shown in Fig. 4. These LCL161-based SNIPERs showed efficient protein knockdown activities against target proteins at nanomolar concentrations [8, 9, 14]. Target protein knockdown activity of SNIPERs was evaluated by immunoblot analysis (Figs. 5a–c, and 6). SNIPER also induces degradation of IAPs (see Note 1). In-depth protocols for immunoblotting are described below. The ERα protein is activated by estrogens and is constantly degraded by the UPS in the cells cultured in normal culture media supplemented with 10% FBS. When ERα protein knockdown activity is obscure due to estrogen-induced downregulation of ERα, cells are depleted of estrogens prior to treatment with SNIPER(ER)s in some experiments. To eliminate estrogens derived from FBS, cells are

SNIPERs Anti-Cancer Agent Methods

333

Ubiquitylation

E2 Ub Ub

E3 ligase

Ub

IAPs Degradation

Target protein

SNIPER

Fig. 1 Schema for SNIPER-induced target protein degradation

Fig. 2 Synthesis schema for SNIPER(ER). Reagents and conditions: (a) 7, EDC, HOBt·H2O, DIPEA, THF, rt, 76%; (b) PPh3, H2O, THF, rt, 94%; (c) 8, EDC, HOBt·H2O, DIPEA, THF, rt, 69% for 3, 93% for 5; (d) 4 M HCl in CPME, MeOH, rt, 82% for 4, 79% for 6; (e) 14, EDC, HOBt·H2O, DIPEA, THF, rt., 70%; (f) p-TsCl, pyridine, 0
C to rt, 73%; (g) 15, K2CO3, DMF, 60
C, 77%; and (h) 4 M HCl in CPME, THF, rt, 32%

precultured for more than 4 days in media supplemented with estrogen-depleted serum (see Note 2). In this condition, ERα protein expression levels are increased due to the lack of estrogens, and the SNIPER(ER)-induced degradation of ERα can be evaluated more easily. Ubiquitylation of ERα is clearly visualized by transfection of HA–ubiquitin plasmid for 24 h (Fig. 6). The addition of a deubiquitylase inhibitor, PR619, along with the

334

Yoshinori Tsukumo et al.

Fig. 3 Synthesis schema for SNIPER(ABL). Reagents and conditions: (a) 7, EDC, HOBt·H2O, DMF, rt, 87%; (b) PPh3, H2O, THF, rt, quant.; (c) 23, HATU, DIPEA, DMF, rt, 81%; (d) 2 M dimethylamine in MeOH, rt, 86%; (e) TFA, rt, 75%; (f) EDC, HOBt·H2O, DMF, rt, quant.; and (g) TFA, rt, 75%

Fig. 4 Synthesis schema for SNIPER(BRD4)-1 and SNIPER(PDE4)-9. Reagents and conditions: (a) 31, HATU, DIPEA, DMF, rt, 50% for 26, 68% for 29 and (b) 6 M aq. HCl, rt, 63% for 28, 68% for 30

proteasome inhibitor, MG132, also improves detection of ubiquitylation-targeted proteins. The mixture of each ligand (4-OHT and LCL161 here) was used as a negative control. Various assays can be used to evaluate antitumor activity in vitro and in vivo. The methods used in our laboratory are described below. Crystal violet staining was used to determine the number of adhesive and living cells (Fig. 7a). WST-8 is a water-soluble

SNIPERs Anti-Cancer Agent Methods

335

Fig. 5 Target protein degradation induced by SNIPER(ER)-87 (Figures were reproduced from reference [8]). (a, b) Protein knockdown activity of SNIPERs. MCF-7 cells were treated with the indicated concentrations of each SNIPER for 6 h. Whole-cell lysates were analyzed by western blotting with the indicated antibodies. Numbers below the ERα panel represent the target protein/actin ratio normalized by vehicle control as 100. Bar graph: mean  S.D. of three independent experiments.*P < 0.05 compared with vehicle control. (c) MCF-7 cells were treated with 100 nM SNIPER(ER)-87 or mixture of the LCL161 derivative and 4-OHT, with or without 10 μM MG132, for 6 h. Cell lysates were analyzed by western blotting

tetrazolium salt and is reduced by mitochondrial NADPH dehydrogenase activity in cells to produce a yellow formazan dye, the intensity of which is directly proportional to the number of living cells. ERα-positive breast tumor cells (MCF-7, T47D) and ERα-negative cells (MDA-MB-231) were used to ensure the estrogen dependency of the growth (Fig. 7a). For in vivo analysis, details are described below. We used an MCF-7 tumor xenograft mouse model to evaluate the in vivo therapeutic efficacy (Fig. 7b, c). After cell inoculation, subcutaneous β-estradiol injection is needed to promote MCF-7 tumor formation in vivo. Protocols for these assays are described in this chapter.

2

Materials

2.1 Design and Synthesis of SNIPERs

1. Reagents for the synthesis of compounds, including SNIPER (ER), SNIPER(ABL), SNIPER(BRD4), and SNIPER(PDE4):

336

Yoshinori Tsukumo et al.

Fig. 6 SNIPER(ER)-87-induced ubiquitylation of ERα protein (This figure was reproduced from Ref. 8). Ubiquitylation of ERα by SNIPER(ER)-87. MCF-7 cells transfected with HA-ubiquitin were treated with the indicated compounds in the presence of 10 μM MG132 for 3 h. Whole-cell lysates (lower panels) and lysates immunoprecipitated with antiHA antibody (upper panel) were analyzed by western blotting with the indicated antibodies. IP immunoprecipitation, IB immunoblot

((E/Z)-4-(1-{4-[2-(methylamino)ethoxy]phenyl}-2-phenyl-1-buten-1-yl)-phenol) ((E/Z)-endoxifen), 2-{2[2-(azidomethoxy)ethoxy]ethoxy}acetic acid, {2-[2-(2-hydroxyethoxy)ethoxy]ethoxy}acetic acid, tetraethylene glycol di( ptoluenesulfonate), p-toluenesulfornyl chloride ( p-TsCl), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (EDC) hydrochloride, 1-hydroxybenzotriazole (HOBt), N,N-diisopropylethylamine (DIPEA), trifluoroacetic acid (TFA), dichloromethane (CH2Cl2), triphenylphosphine (PPh3), methanol (MeOH), ethanol (EtOH), n-hexane, ethyl acetate (EtOAc), N,N-dimethylformamide (DMF), sodium hydrogen carbonate (NaHCO3), potassium carbonate (K2CO3), ammonium chloride (NH4Cl), magnesium sulfate (MgSO4), sodium sulfate (Na2SO4), hydrochloric acid (HCl), ammonium hydroxide (aq. NH3), tetrahydrofuran (THF), cyclopentyl methyl ether (CPME), and 2 M methanolic dimethylamine. 2. Silica gel 60 N (neutral), spherical, particle size 40–100 μm. 3. NH silica gel. 4. N-Boc-bestatin [5].

SNIPERs Anti-Cancer Agent Methods

337

Fig. 7 Antitumor activity of SNIPER(ER)-87 (Figures were reproduced from Ref. 8.) (a) Growth inhibition of ERα-positive human breast tumor cells by SNIPER(ER)-87. Cells were treated with the indicated concentrations of SNIPER(ER)-87 for 72 h, and the cell growth was evaluated by cell viability assay. The data represent mean  S.D. (n ¼ 3). (b) ERα protein levels in tumor xenografts were analyzed by western blotting; Bar graphs: mean  S.D. of each group (n ¼ 12). *, P < 0.001 (two-sided Student’s t test). (c) SNIPER(ER)-87 inhibits the growth of MCF-7 orthotopic breast tumor xenografts in nude mice. Tumor volume: mean  S.D. of each group (mice: n ¼ 9 each; tumors: n ¼ 18 each) *P < 0.0001 (two-sided Student’s t test). Mice were administered vehicle or SNIPER(ER)-87 (30 mg/kg, intraperitoneally, every 24 h)

5. N-Boc-MV1 [5]. 6. SEM-protected (E/Z)-Endoxifen [8]. 7. N-Boc-LCL161-OH [8]. 8. N-Fmoc-bestatin [14]. 9. Dasatinib derivative [16]. 10. N-Boc-LCL161-PEG3-CO2H [14]. 11. (+)-JQ1 derivative [16]. 12. PDE4 inhibitor [17]. 13. N-Boc-LCL161-PEG3-NH2 [8]. 2.2 Target Protein Degradation

1. Cells: MCF-7, T47D, ZR-75-1, and MDA-MB-231 human breast carcinoma cell lines. 2. RPMI 1640 medium: RPMI 1640, 10% heat-inactivated fetal bovine serum (FBS), and 100 μg/mL of kanamycin (used for MCF-7, T47D, ZR-75-1, MDA-MB-231, K562, and LNCaP cells). 3. Phenol red-free RPMI1640 medium. 4. Charcoal/dextran-treated FBS (GE Health care, Little Chalfont, United Kingdom).

338

Yoshinori Tsukumo et al.

5. Six-well cell culture plates. 6. Synthesized compounds: 10 mM SNIPER(ER)s (compound14, 19, and 87,) in DMSO stock solution. 7. 10–30 μM 4-Hydroxytamoxifen (4-OHT). 8. 10–30 μM Bestatin methyl ester (MeBS, Millipore, Darmstadt, Germany). 9. 10 μM MG-132 (proteasome inhibitor, Z-Leu-Leu-Leu-H [aldehyde]) (Peptide Institute, Osaka, Japan). 10. 500 μM-2 mM Hydrogen peroxide. 11. Phosphate-buffered saline (PBS). 12. Lysis buffer: 0.1 M Tris–HCl (pH 7), 1% SDS, 10% glycerol. 13. 96-well plate. 14. BCA Protein Assay Reagent Kit. 15. Microplate reader. 16. SDS loading buffer (5): 0.5 M Tris–HCl, pH 7, 5% SDS, 50% glycerol, 0.5 M dithiothreitol, and 0.05% bromophenol blue. 17. Prestained protein marker, broad range (7–175 kDa). 18. Precast gel: Perfect NT Gel A (10–15%, 18-well) (DRC, Tokyo, Japan). 19. SDS-polyacrylamide gel electrophoresis (PAGE) apparatus. 20. SDS-PAGE running buffer: 25 mM Tris, 192 mM Glycine, 0.1% SDS. 21. Filter paper. 22. Polyvinylidene dione (PVDF) membrane. 23. Transfer buffer: 25 mM Tris, 192 mM glycine, 20% (v/v) methanol. 24. Transfer apparatus for western blotting, including power supply. 25. Tris buffered saline (TBS): 50 mM Tris–HCl, pH 7.5, 150 mM NaCl, with 0.05% Tween-20 (TBS-T). 26. Blocking solution: 5% nonfat dried skim milk in TBS-T. 27. Antibodies: antiERα rabbit monoclonal antibody (Cell Signaling Technology, Danvers, MA); antiERα rabbit polyclonal antibody (Santa Cruz Biotechnology, Dallas, TX); anticIAP1 goat pAb (R&D Systems, Minneapolis, MN); anticIAP1 rat mAb (Enzo Life Sciences, Farmingdale, NY); antiβ-actin mouse mAb (Sigma); antiXIAP rabbit pAb (Cell Signaling Technology); antirabbit IgG HRP conjugate, antimouse IgG HRP conjugate. 28. Chemiluminescent substrate for HRP. 29. LAS-3000 lumino-image analyzer (Fuji, Tokyo, Japan).

SNIPERs Anti-Cancer Agent Methods

2.3

Cell Death

339

1. 96-well plate. 2. 10% crystal violet in methanol. 3. 1% SDS solution. 4. Microplate reader.

2.4 Ubiquitylation Assay

1. Lipofectamine Scientific).

2000

and

Opti-MEM

(Thermo

Fisher

2. pcDNA3-HA-ubiquitin plasmid. 3. 10 μM MG132. 4. Lysis buffer: 0.1 M Tris–HCl (pH 7), 1% SDS, 10% glycerol. 5. Dilution buffer: 0.1 M Tris-HCl (pH 7.5). 6. AntiHA agarose-conjugated beads (Sigma). 2.5 In Vivo Protein Knockdown

1. BALB/c nude mice (Clea Japan, Tokyo, Japan). 2. Matrigel (Corning Life Sciences). 3. β-estradiol (Innovative Research of America, Sarasota, FL.

2.6 In Vivo Tumor Growth Inhibition

1. BALB/c nude mice. 2. Matrigel (Corning Life Sciences). 3. β-estradiol. 4. Caliper.

3

Methods

3.1 Design and Synthesis of SNIPER (ER)-14 and -19

1. Stir a mixture of 9.03 mmol of (E/Z)-endoxifen, 27.1 mmol of DIPEA, 10.8 mmol of HOBt·H2O, 10.8 mmol of EDC, and 9.03 mmol of 2-{2-[2-(azidomethoxy)ethoxy]ethoxy}acetic acid (7) into 80 mL of THF at room temperature overnight. 2. Pour water into the reaction mixture and extract with EtOAc. 3. Wash the organic extract with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on silica gel (0–20% MeOH in EtOAc) to give 4.06 g of Compound 1 (76%). 4. Stir a mixture of 6.90 mmol of 1, 13.8 mmol of PPh3 in 80 mL of THF and 20 mL of H2O at room temperature overnight. After removing the solvent, purify the residue by column chromatography on NH silica gel (eluted with 0–20% MeOH in EtOAc) to give 2 (94%). 5. Stir a mixture of 0.510 mmol of 2, 2.5 mmol of DIPEA (0.44 mL), 0.607 mmol of HOBt·H2O, 0.61 mmol of EDC, and 0.510 mmol of N-Boc bestatin (8) [5] into 4 mL of THF at room temperature overnight.

340

Yoshinori Tsukumo et al.

6. Pour water into the reaction mixture and extract with EtOAc. 7. Wash the organic extract with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–20% MeOH in EtOAc) to give 3 (69%). 8. Stir a mixture of 0.100 mmol of 3 in 4 M HCl in 3.2 mmol of CPME and 1 mL of MeOH at room temperature for 4 h. After removing the volatiles, neutralize the residue with saturated aq. NaHCO3 and extract with EtOAc–THF. 9. Wash the organic extracts with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–50% MeOH in EtOAc) to give 4 (SNIPER(ER)-14) (82%, E/Z ¼ 1:1, chiral HPLC). Compound 6 (SNIPER(ER)-19) is prepared from Compounds 2 and 9 in a similar manner, via 5, to that described for Compound 4 and obtained as a colorless oil (79%, E/Z ¼ 1:1, chiral HPLC). 3.2 Design and Synthesis of SNIPER (ER)-87

1. Stir a mixture of 0.695 mmol of 10 (SEM-protected (E/Z)endoxifen) [8], 0.835 mmol of {2-[2-(2-hydroxyethoxy)ethoxy]ethoxy}acetic acid (14), 1.04 mmol of HOBt·H2O, 2.1 mmol of DIPEA, and 1.0 mmol of EDC into 10 mL of THF at room temperature overnight. 2. Pour water into the mixture and extract with EtOAc. 3. Wash the organic extracts with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–30% MeOH in EtOAc) to give 11 (70%). 4. Add 1.94 mmol of p-TsCl to a solution of 0.486 mmol of 11 in 5 mL of pyridine at 0
C and stir at room temperature for 6 h. After removing the solvent, dilute with EtOAc. 5. Pour saturated aq. NH4Cl into the solution and extract with EtOAc. 6. Wash the organic layer with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on silica gel (0–30% MeOH in EtOAc) to give the O-mesylated form of 11 (73%). 7. Stir a mixture of the O-mesylated form of 0.410 mmol of 11, 0.411 mmol of Boc-LCL161 derivative 15 [8], and 0.492 mmol of K2CO3 in 3 mL of DMF at 60
C overnight. 8. Pour saturated aq. NH4Cl into the reaction mixture and extract with EtOAc.

SNIPERs Anti-Cancer Agent Methods

341

9. Wash the organic extracts with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0%–30% MeOH in EtOAc) to give 12 (77%). 10. Stir a mixture of 0.267 mmol of 12 in 1 mL of THF and 4 M HCl in 13.2 mmol of CPME (3.3 mL) at room temperature for 2 h. After removing the volatiles, neutralize the residue with saturated aq. NaHCO3 and extract with EtOAc. 11. Wash with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–20% MeOH in EtOAc) and preparative HPLC to give 90 mg of 13 (SNIPER(ER)-87;32%, E/Z ¼ 1:1, chiral HPLC). 3.3 Design and Synthesis of SNIPER (ABL)-20 and -19

1. Stir a mixture of 0.80 mmol of dasatinib derivative 16 [15], 1.12 mmol of 2-{2-[2-(2-azidoethoxy)ethoxy]ethoxy}acetic acid (7), 1.12 mmol of HOBt·H2O, and 1.1 mmol of EDC in 6 mL of DMF at room temperature overnight. 2. Dilute the reaction mixture with EtOAc, wash with 5% aq. NaHCO3 and brine, and dry over Na2SO4. After removing the solvent, purify the residue by column chromatography on silica gel (0–20%MeOH in EtOAc) to give 160 (87%). 3. Stir a mixture of 0.69 mmol of 160 and 1.38 mmol of PPh3 in 6 mL of THF and 0.2 mL of water at room temperature overnight. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–40% MeOH in EtOAc) to give the corresponding amine (quant). 4. Stir a mixture of 0.36 mmol of the obtained amine, 0.43 mmol of Fmoc-bestatin (23) [6] or Boc-MV1 (9) [5], 0.51 mmol of HATU, and 0.75 mmol of DIPEA in DMF at 0
C for 10 min. 5. Dilute the reaction mixture with EtOAc, wash with 5% aq. NaHCO and brine, and dry over Na2SO4. After removing solvent, purify the residue by column chromatography on silica gel (0%–20%MeOH in EtOAc) to give 17 (81%) and 19 (81%), respectively. 6. Stir a mixture of 0.29 mmol of 17 and 2 M dimethylamine in MeOH at room temperature overnight. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–20% MeOH in EtOAc) to give 18 (SNIPER (ABL)-20) (86%). 7. Stir a mixture of 0.29 mmol of 19 in 6 mL of TFA at room temperature for 10 min. Dilute the reaction mixture with 5 mL of toluene and concentrate in vacuo. 8. Dissolve the residue in EtOAc–THF (ca 3:1), wash with saturated aq. NaHCO3 and brine, and dry over Na2SO4.

342

Yoshinori Tsukumo et al.

After removing the solvent, purify the residue by column chromatography on NH silica gel (0–20% MeOH in EtOAc) to give 20 (SNIPER(ABL)-19; 92%). 3.4 Design and Synthesis of SNIPER (ABL)-39

1. Stir a mixture of 0.20 mmol of dasatinib derivative 16 (89 mg), LCL161-PEG3-CO2H [14], 0.20 mmol of 24, 0.28 mmol of HOBt·H2O, and 0.28 mmol of EDC in 1.5 mL of DMF at room temperature overnight. 2. Dilute the reaction mixture with EtOAc, wash with 5% aq. NaHCO3 and brine, and dry over Na2SO4. After removing the solvent, purify the residue by column chromatography on silica gel (0–40% MeOH in EtOAc) to give 21 (quant.). 3. Stir a mixture of 0.20 mmol of 21 in 5 mL of TFA at room temperature for 10 min. 4. Dilute the mixture with 5 mL of toluene and concentrate in vacuo. 5. Dilute the residue with EtOAc, wash with 5% aq. Na2CO3 and brine, dry over Na2SO4, and concentrate in vacuo. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–30% MeOH in EtOAc) to give 22 (SNIPER(ABL)-39; 75%).

3.5 Design and Synthesis of SNIPER (BRD4)-1 and SNIPER (PDE4)-9

1. Stir a mixture of 0.348 mmol of (+)-JQ1 derivative 25 [16] or PDE4 inhibitor 28 [17] and 0.74 mmol of DIPEA, 0.736 mmol of HATU, and 0.368 mmol of N-Boc LCL161PEG3-NH2 31 [8] in 3 mL of MeCN at room temperature for 1 h. 2. Pass the mixture through an NH silica gel pad and elute with THF. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–5% MeOH in EtOAc) to give 26 (50% in two steps) or 29 (68% in two steps). 3. Stir a mixture of 0.182 mmol of 26 or 29 and 6 M aq. HCl (12.0 mmol) in 4 mL of THF at room temperature for 5 h. Neutralize the mixture with saturated aq. NaHCO3 and extract with EtOAc. 4. Wash the organic layer with water and brine and dry over MgSO4. After removing the solvent, purify the residue by column chromatography on NH silica gel (0–10% MeOH in EtOAc or 10–30% MeOH in EtOAc) to give 28 SNIPER (BRD4)-1 (63%) or 30 (SNIPER(PDE4)-9; 68%), respectively.

3.6 Analysis of Target Protein Degradation by Immunoblot

1. Maintain MCF-7, T47D, ZR-75-1, and MDA-MB-231 cells in RPMI 1640 medium. 2. Plate cells at a density of 4  105 cells/well in 6-well cell culture plates. To prepare for experiments under estrogen-depleted

SNIPERs Anti-Cancer Agent Methods

343

conditions, switch the culture medium to phenol red-free medium containing 4% charcoal/dextran-treated FBS and kanamycin for more than 72 h. Then, change the medium to the cell-appropriate medium that contains 0.2% charcoal/dextran-treated FBS and kanamycin for 24 h (see Note 2). 3. Treat the cells with 10 μM 4-hydroxytamoxifen or SNIPER (ER)-14, 19, and 87 for 6 h in the absence and the presence of 10 μM proteasome inhibitor MG-132. When MG-132 is used to block the degradation of ERα by SNIPER(ER), treat the cells with MG-132 for 30 min before adding the SNIPER(ER). 4. After incubation, aspirate the medium, add 1 mL/well of cold PBS, scrape adherent cells from each well with a cell scraper, and transfer the cells to 1.5-mL tubes. 5. Centrifuge the cells at 800  g for 5 min. 6. Aspirate the PBS from the pellet and add 100 μL of lysis buffer. 7. Vortex the sample to dissolve the cell pellet. 8. Boil the sample at 95
C for 10 min. 9. Place the cell lysate on ice for 2 min. 10. Centrifuge the cell lysate at 8000  g at 4
C for 10 min to remove undissolved debris. 11. Transfer the supernatant (~90 μL) to a new 1.5-mL tube. 12. Determine protein concentrations by the BCA method [14] using the BCA assay kit. Add diluted cell lysate or BSA standard and working reagent for BCA assay (50:1 mixture of Solution A and B) to each well in a 96-well plate, mix briefly on a plate shaker, and incubate at 37
C for 30 min. Read the absorbance at 550–570 nm, using a microplate reader (see Notes 3 and 4). 13. To prepare samples for SDS-PAGE, dilute equal amounts of protein from each cell lysate to equal volumes with lysis buffer. Add 5 SDS sample buffer and boil the samples at 95
C for 5 min. 14. Load the samples onto a polyacrylamide gel and electrophorese for 0.5–1 h at 200 V in SDS-PAGE running buffer. 15. Transfer the proteins electrophoretically onto a PVDF membrane for 10–24 h at 50–120 V (1200 Vh) in transfer buffer (see Note 5). 16. Block the membranes in blocking solution for 1 h on a shaker. 17. Immunoblot the membranes with the primary antibody (for ERα or β-actin) for 2 h at room temperature in blocking solution (see Note 6). 18. Remove the primary antibodies. 19. Wash the PVDF membranes three times with TBS-T for 5 min each.

344

Yoshinori Tsukumo et al.

20. Incubate the membranes with the respective secondary antibodies conjugated with horseradish peroxidase for 1 h at room temperature in blocking solution. 21. Remove the secondary antibodies. 22. Wash the PVDF membranes three times with TBS-T for 5 min each. 23. Visualize the protein signals using chemiluminescent substrate and the appropriate imaging system (Fig. 5a). 24. Quantify the light emission with an LAS-3000 lumino-image analyzer and represent the ERα/β-actin ratio, normalized to vehicle control set at 100 (Fig. 5b). 3.7 Analysis of Ubiquitylation

1. Maintain MCF-7 cells in RPMI 1640 medium. 2. Transfect with pcDNA3-HA-ubiquitin for 24 h. 3. Incubate the transfected cells with the indicated compounds in the presence of 10 μM MG132 for 3 h. 4. Harvest cells and lyse in SDS lysis buffer. 5. Boil the sample at 95
C for 10 min and dilute ten times with 0.1 M Tris–HCl (pH 7.5). 6. Immunoprecipitate with antiHA agarose-conjugated beads (Sigma) at 4
C for 2 h. 7. Wash the precipitates with wash buffer four times.

3.8 Analysis of Cell Death by Crystal Violet Staining

Crystal violet staining is used to determine the number of adhesive and living cells. 1. Plate the cells at a density of 1.5  104 cells/well in 96-well plates. 2. The next day, treat the cells with the indicated concentrations of SNIPER(ER)-87 for 72 h. 3. Aspirate the medium and add 50 μL of PBS containing 0.1% crystal violet. 4. Incubate the cells on a gentle shaker at room temperature for 15 min. 5. Wash the cells thoroughly with distilled water. 6. Dry the cells at 37
C for 2 h. 7. Add 100 μL of 1% SDS solution. 8. Incubate the cells on a shaker at room temperature for 1 h. 9. Read the absorbance at 600 nm, using a microplate reader (Fig. 7a).

SNIPERs Anti-Cancer Agent Methods

3.9 Analysis of In Vivo Protein Knockdown

345

1. Keep mice in pathogen-free animal facilities with 12-h light/ dark cycles. Feed them rodent chow and water ad libitum, according to the guidelines approved by your institute. 2. Mix suspension of 1  107 MCF-7 cells with an equal volume of Matrigel and inject the suspension (100 μL total) into the left and right mammary fat pads of 6-week-old female BALB/c nude mice. 3. Inject subcutaneously β-estradiol solution into the neck twice at intervals of 6 days. 4. Fourteen days after the last β-estradiol injection, administer vehicle or 30 mg/kg SNIPER(ER)-87 into the mice intraperitoneally. 5. After 24 h, excise the tissues from the sacrificed mice.

3.10 Analysis of In Vivo Tumor Growth Inhibition

1. Keep mice in pathogen-free animal facilities with 12-h light/ dark cycles. Feed them rodent chow and water ad libitum according to the guidelines approved by your institute. 2. Mix a suspension of 1  107 MCF-7 cells with an equal volume of Matrigel. 3. Inoculate the suspension (100 μL total) into the left and right mammary fat pads of 6-week-old female BALB/c nude mice that had received β-estradiol pellets (6 μg per day) under the neck skin. 4. After 4 days, randomize and divide the mice with ~100-mm3 tumors into two groups (n ¼ 9). 5. Treat one group with vehicle as a control for the dosing vehicle. Administer SNIPER(ER)-87 (30 mg/kg, intraperitoneally, every 24 h) into the other group. 6. Measure the tumor volumes every 2 days using a caliper. Calculate the volume according to the standard formula: (length  width2)/2. 7. At 2 weeks, excise the tumors from sacrificed mice.

4

Notes 1. SNIPER induces degradation of IAP family proteins simultaneously with target proteins. cIAP1 is autoubiquitylated and degraded by binding of SNIPERs, while XIAP requires ternary complex formation, XIAP-SNIPER-target proteins, for degradation. 2. On binding with estrogen, ERα translocates to the nucleus and activates gene expression; it is then degraded via the ubiquitinproteasome system. To differentiate SNIPER-induced ERα

346

Yoshinori Tsukumo et al.

degradation from physiological estrogen degradation, cells are cultured in a medium that contains estrogen-depleted (charcoal/dextran-treated) serum. Phenol red-free medium is also used, because phenol red mimics estrogenic activity. 3. The BCA protein assay is compatible with the detergents used in the experiment. Absorbance of 1% SDS lysis buffer reacting with the BCA assay reagents is negligible. 4. According to the instructions, reaction complexes exhibit absorption maximum at 562 nm. Absorbance measurement in this assay is compatible between 540 nm and 590 nm. 5. The PVDF membrane is immersed in 100% methanol and then transferred into transfer buffer before use. 6. Alternatively, CanGet Signal (Toyobo, Osaka, Japan) or similar reagents can be used to enhance signals detected by antibodies.

Acknowledgments This study was supported by MEXT/JSPS KAKENHI (grants JP19K07724 and JP16K18444 to Y.T., JP19K16333 to G.T., JP18K07311 to N.S., JP18K06567 to N.O., JP17K08385 to Y. D., JP16H05090, and JP18H05502 to M.N.) and by AMED (grants JP17cm0106522j0002 to N.S., JP19cm0106136, JP19ak0101073 to N.O., JP19ak0101073j0603, and JP19im0210616j0002 to M.N.). We are grateful to Journal of Biological Chemistry and the American Society for Biochemistry and Molecular Biology (ASBMB) for allowing the reproduction of Figs. 5, 6, and 7, which were originally published in the Journal of Biological Chemistry [8] (Ohoka N, Okuhira K, Ito M et al. In vivo knockdown of pathogenic proteins via specific and nongenetic inhibitor of apoptosis protein (IAP)-dependent protein erasers (SNIPERs). J Biol Chem. 2017; 11:4556–4570. © the American Society for Biochemistry and Molecular Biology.) We also thank Marla Brunker, from Edanz Group (www.edanzediting.com/ac), for editing a draft of this manuscript. References 1. Okuhira K, Ohoka N, Sai K et al (2011) Specific degradation of CRABP-II via cIAP1mediated ubiquitylation induced by hybrid molecules that crosslink cIAP1 and the target protein. FEBS Lett 585:1147–1152 2. Itoh Y, Ishikawa M, Naito M et al (2010) Protein knockdown using methyl bestatinligand hybrid molecules: design and synthesis of inducers of ubiquitination-mediated

degradation of cellular retinoic acid-binding proteins. J Am Chem Soc 132:5820–5826 3. Itoh Y, Ishikawa M, Kitaguchi R et al (2011) Development of target protein-selective degradation inducer for protein knockdown. Bioorg Med Chem 19:3229–3241 4. Itoh Y, Kitaguchi R, Ishikawa M et al (2011) Design, synthesis and biological evaluation of nuclear receptor-degradation inducers. Bioorg Med Chem 19:6768–6778

SNIPERs Anti-Cancer Agent Methods 5. Itoh Y, Ishikawa M, Kitaguchi R et al (2012) Double protein knockdown of cIAP1 and CRABP-II using a hybrid molecule consisting of ATRA and IAPs antagonist. Bioorg Med Chem Lett 22:4453–4457 6. Ohoka N, Nagai K, Hattori T et al (2014) Cancer cell death induced by novel small molecules degrading the TACC3 protein via the ubiquitin-proteasome pathway. Cell Death Dis 5:e1513 7. Okuhira K, Demizu Y, Hattori T et al (2013) Development of hybrid small molecules that induce degradation of estrogen receptor-alpha and necrotic cell death in breast cancer cells. Cancer Sci 11:1492–1498 8. Ohoka N, Okuhira K, Ito M et al (2017) In vivo knockdown of pathogenic proteins via specific and nongenetic inhibitor of apoptosis protein (IAP)-dependent protein erasers (SNIPERs). J Biol Chem 11:4556–4570 9. Ohoka N, Morita Y, Nagai K et al (2018) Derivatization of inhibitor of apoptosis protein (IAP) ligands yields improved inducers of estrogen receptor α degradation. J Biol Chem 18:6776–6790 10. Okuhira K, Demizu Y, Hattori T et al (2016) Molecular design, synthesis, and evaluation of SNIPER(ER) that induces proteasomal degradation of ERα. Methods Mol Biol 1366:549–560

347

11. Saito M, Aoyagi T, Umezawa H et al (1976) Bestatin, a new specific inhibitor of aminopeptidases, enhances activation of small lymphocytes by concanavalin A. Biochem Biophys Res Commun 76:526–533 12. Umezawa H, Aoyagi T, Suda H et al (1976) Bestatin, an inhibitor of aminopeptidase B, produced by actinomycetes. J Antibiot 29:97–99 13. Sekine K, Takubo K, Kikuchi R et al (2008) Small molecules destabilize cIAP1 by activating auto-ubiquitylation. J Biol Chem 283:8961–8968 14. Shibata N, Miyamoto N, Nagai K et al (2017) Development of protein degradation inducers of oncogenic BCR- ABL protein by conjugation of ABL kinase inhibitors and IAP ligands. Cancer Sci 108:1657–1666 15. Veach DR, Namavari M, Pillarsetty N et al (2007) Synthesis and biological evaluation of a fluorine-18 derivative of dasatinib. J Med Chem 50:5853–5857 16. Winter GE, Buckley DL, Paulk J et al (2015) Drug development. Phthalimide conjugation as a strategy for in vivo target protein degradation. Science 348:1376–1381 17. Govek SP, Oshiro G, Anzola JV et al (2010) Water-soluble PDE4 inhibitors for the treatment of dry eye. Bioorg Med Chem Lett 20:2928–2932

INDEX A

E

Affinity intracellular ......................................97, 274, 275, 280 proteins .................................... 17, 19, 21, 24, 60, 61, 70, 72–74, 80, 82, 84, 85, 92, 97, 99, 116, 117, 123, 127, 138, 146, 234, 269 Affinity beads................................................60, 61, 69–74

E3 ubiquitin ligase ..................................... 107, 116, 118, 135, 177, 217, 315, 331

F FunFOLD3 ...............................................................43–56

H

B Biacore .....................................................................6–9, 20 Binding affinities ........................................ 16, 17, 81, 82, 84–87, 89, 92, 99, 117, 123, 127, 128, 275, 279, 286, 297 Binding kinetics................................................................. 6 Biophysical methods ............................... 4, 88, 89, 91–98 Biotin ............................... 60–64, 66, 68, 71–74, 97, 193

Heterobifunctional degraders ..................................3, 4, 6 High-throughput assays .......................................... 35, 88, 89, 136, 137, 266, 284, 292 detection .............................................................. 21–40 Hook effect.................................... 80–83, 89, 90, 92, 93, 103, 104, 131, 148, 248, 326

I

C Cancer..................................... 83, 87, 97, 104, 136, 149, 176, 203, 284, 332 Cellular thermal shift assay (CETSA) .....................21–28, 31, 33–36, 38–40, 99 CELMoDs ............................................................ 283–299 Cereblon (CRBN)..................................7, 60, 61, 70, 71, 80, 82, 89, 99, 135, 136, 140, 153–156, 162, 163, 165, 168, 249, 266, 268, 269, 271–274, 276–280, 283–299, 316, 326 Cooperativities ........................80–85, 87, 89, 92, 95, 96, 107, 117, 118, 128, 129, 131 Critical assessment of techniques for protein structure prediction (CASP) ..................... 46, 48, 51–53, 55 Crystal structures ......................................................79–88

D Degraders .......................................59–61, 83–85, 87, 89, 92, 97, 99, 101, 104, 105, 116, 117, 136, 152, 153, 176, 181, 182, 200, 201, 247–252, 265, 279 Docking ................................................................... 46, 52, 54–56, 283

IMiDs......................................................6, 80, 82, 83, 89, 135, 249, 252 Interacting partners FunFOLD3 ......................................................... 43–56 protein–ligand binding site prediction methods ......................................................... 43–56 Isobaric labeling isobaric tag............................................................... 302 Isothermal calorimetry (ITC) ..........................84, 89, 92, 93, 96

K Kinetics ............................. 5, 6, 9, 16, 17, 19, 84, 87, 92, 96–99, 103, 105, 107, 152–156, 161, 162, 164–165, 167, 242, 248, 249, 252

L LC-MS/MS....................................... 204, 205, 207, 212, 301–303, 306, 309, 310 Ligand binding site prediction methods .................43–56 Ligand evaluation using affinity beads and cell extracts .................. 59–75 for ubiquitin ligases............................................. 59–75

Angela M. Cacace et al. (eds.), Targeted Protein Degradation: Methods and Protocols, Methods in Molecular Biology, vol. 2365, https://doi.org/10.1007/978-1-0716-1665-9, © Springer Science+Business Media, LLC, part of Springer Nature 2021

349

TARGETED PROTEIN DEGRADATION: METHODS AND PROTOCOLS

350 Index

Ligand interactions using surface plasmon resonance (SPR) biacore ......................................................................... 6 binding affinity ....................................................16, 17 binding kinetics .....................................................5, 17 cereblon ....................................................................... 7 Ligand-protein binding using a SplitLuc cellular thermal shift assay (CETSA) high-throughput screening ...................................... 22 ligand binding ..................................................... 21–40 luciferase ...................................................... 22, 24, 27, 33, 38 target engagement .............................................21, 22, 35, 36, 39 thermal stability...................................................21, 24 Liquid chromatography (LC) LC-MS/MS............................................301–303, 306

M Mass spectrometry .......................... 72, 97, 99, 105, 187, 203–214, 284 MicroScale thermophoresis (MST)..................... 115–132 Modelling ........................................................... 46, 84, 87

N NanoBRET................................................... 99, 101, 103, 151–168, 265–282 Neosubstrates ....................................................... 283–299

O Optical control ..................................................... 315–327

P Permeability..........................................97, 161, 265, 266, 274, 279, 281, 326 PHOTACs photopharmacology ................................................ 316 photoswitches................................................. 316, 319 Polyubiquitin chains (chain selectivity) ......................203, 217, 218 Pomalidomide .....................................6, 7, 9, 19, 83, 284 Post-translational modification ............................. 36, 175 Proteasomal unfolding......................................... 217–243 Proteasomes...............................101, 116, 152, 156, 161, 162, 176, 180, 183, 185, 186, 197–199, 201, 203, 217–219, 223, 224, 233, 234, 236–240, 242, 247, 285, 315, 333, 338, 343 Protein abundance ............................................... 301–311 Protein degradation ......... 24, 59, 80, 97, 103–107, 136, 175–183, 247–263, 301, 315–327, 331, 333, 335, 336, 338, 342 Proteolysis........................................................24, 74, 116, 185, 315

Proteolysis-targeting chimeras (PROTACs) .........................................24, 79–108, 115–132, 135–149, 151–155, 185, 249, 252, 265–267, 272–274, 277, 279, 315–317, 319, 326 Proteome profiling............................................... 301–311 Proteomics............................................61, 86, 87, 97, 99, 104, 105, 204 PTMScan ................................... 204–206, 209, 211, 213, 214, 302–304 Pull-down Assay ................................................... 135–149

S SNIPERs............................................................... 331–346 SplitLuc cellular thermal shift assay (CETSA) ........................................................21–41 Streptavidin................................ 7, 19, 60–63, 66, 69–71, 73, 74, 89, 99, 194, 195 Surface plasmon resonance (SPR).......................... 12, 82, 96–98, 137, 146 Synthesis .......................30, 61–64, 67, 68, 87, 284, 315, 331–336, 338, 340–343, 345, 346

T Tandem mass spectrometry (TMS)..................... 301–312 Tandem ubiquitin binding entities (TUBEs) ......................................... 26–29, 31, 33, 38, 39, 62, 63, 66, 69, 72, 93, 185–202, 271, 291, 343 Targeted protein degradation (TPD) ............97, 99, 101, 103–106, 108, 135, 136, 145, 283, 301, 316 Target engagement ..............................22, 35, 36, 39, 97, 99, 265–267, 269, 270, 272–274, 276, 277 Temperature related intensity change (TRIC) ...................................................... 115–132 Ternary complexes isothermal calorimetry (ITC) ..................... 79, 80, 83, 86, 90, 97, 101, 104 kinetic detection of E3:PROTAC:target ternary complexes using NanoBRET technology in live cells............................................................ 151–168 pull-down assay of the E3 ligase:PROTAC:substrate ternary complex........................................ 135–149 Thalidomide .............................................. 6, 7, 9, 82, 284 Thermal shift assay (CETSA) ......................................... 99

U Ubiquitin (Ub)............................................. 59, 101, 103, 116, 120, 125, 128, 131, 135, 152, 175, 177, 179, 183, 186, 189, 190, 194, 195, 200, 203, 204, 217, 218, 220, 221, 227, 228, 233–235, 247, 265, 283, 285, 287, 294, 298, 331, 333

TARGETED PROTEIN DEGRADATION: METHODS Ubiquitination.............................................. 3, 80, 96, 97, 101, 103–106, 136, 137, 152, 175–182, 193, 196, 199–201, 203–215, 218–220, 223–225, 227–234, 241, 242, 247, 283, 285, 287, 288, 294–296, 298 Ubiquitin ligases ........................................ 116, 117, 131, 143, 182, 204 Ubiquitin proteasome system (UPS)............................. 59 Ubiquitylation

AND

PROTOCOLS Index 351

mass spectrometry-based analysis of protein ubiquitination using K-ε-GG remnant antibody enrichment................................................ 203–214 tandem ubiquitin binding entities (TUBEs) ................................................... 185–202

V VHL E3 ligase .....................................268, 269, 271–274