Targeting Bromodomains – BET and the wider target class DiscoveRx & SGC Symposium - 12 th September 2013 Matthew Lindon Overview � Brief introduction to epigenetics and bromodomains � Identification of BET inhibitors � Delivering selectivity in the BET family � Beyond BET and target class approaches to other Bromodomain family members 1
Chromatin & Epigenetics DNA is packaged around histones and other proteins to form chromatin Chromatin is highly dynamic material which undergoes remodelling to allow suppression or activation of genes A number of Epigenetic mechanisms control chromatin remodelling including post translational modifications (PTMs) post-translational modifications (PTMs) on histone tails Dysregulation of histone PTMs implicated in human disease Histones: Post-translational modifications PTM of histone tails • PTMs can have a direct impact on physical properties of individual nucleosomes e.g. neutralisation of charge • PTM are recognised by specialised reader domains. 2
Bromodomains bind to acetylated lysine residues Serendipity - BET family BRDs as tractable targets Jonathan Wilde Helen Flynn Lysate from HepG2 Princen Julia White THP1 or HEK293 N 5-10 x 10 8 cells per pulldown N N H N N O O Modified Compound p attached to Reactigel Matrix GW841819X Series Elution with SDS PMM “Black Box” transcriptional Separate PROTEIN Data Analysis on 1-D IDENTITY screen identified a series of ApoA1 Gel DDA A I LC/MS/MS up-regulators (CVU) MS: DBP Some BIAcore Results - GW842819X BZD active N Brd2 67-200 Brd2 338-473 N N O Steady State: Fc=1 Spot=1-r corr Steady State: Fc=1 Spot=1-r corr Steady State: Fc 1 Spot 2 r corr Steady State: Fc=1 Spot=2-r corr U U RU RU 60 RU RU 120 N O N 100 50 40 80 GW841819X 30 60 KD = 2.25e-7M KD < 7.0e-8M 20 40 Rmax = 74 => 70% surface Rmax = 41 => 40% surface is active. 10 is active. 20 koff ~ 0.01s -1 ± 0.03 s -1 0 0 kon -10 -20 1e-9 1e-8 Conc 1e-7 1e-6 1e-5 M 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 Co nc M Fc=1 Spot=1-r corr Fc=1 Spot=2-r corr 60 120 50 100 40 80 30 60 20 40 10 20 0 0 -10 -20 � Target confirmed as the BET family of bromodomains -20 -40 -50 0 50 100 150 200 250 300 -50 0 50 100 150 200 250 300 Chung et al - Journal of Medicinal Chemistry (2011), 54(11), 3827-3838 3
BET Bromodomain Inhibition � BET family made up of Brd2, Brd3, Brd4 and BrdT � Each family member has two distinct bromodomain binding pockets � Inhibitors published to date typically bind all 8 of the BET bromodomains with similar affinity, somewhat limiting their utility as specific probes � Growing body of literature demonstrates Pan-BET inhibition leads to profound preclinical biology � Is it possible to identify selectivity within the BET family? BET Selectivity – Reason to believe (I-BET151) Brd4 BD1 � Very high domain Brd4 BD2 Brd4 BD2 sequence homology would suggest selectivity will be difficult to achieve � Key residues for targeting identified � Chemistry focused on a Asp144 His437 specific vector Tyr97 Tyr390 � How will we know we’ve been successful? 4
Brd4 Bromodomain Mutants � Requirement for robust reagents and assays � Brd4 protein mutations – Brd4 Y390A Brd4 Y390A – Mutation of C-terminal BD2 Bromodomain Mutation of C terminal BD2 Bromodomain – Brd4 Y97A – Mutation of N-terminal BD1 Bromodomain – Brd4 Y97A Y390A – Mutation of both Bromodomains BD1 BD2 C Wild type N C C Y390A Y390A N N C Y97A N C Y97A Y390A N Ligand Based Assays Brd4 Wild Type 400 380 Brd4 Y97A 360 360 Brd4 Y390A 340 320 Brd4 Double Mutant 300 mP 280 [Brd] (M) Kd (nM) Ser (nM) 260 240 Brd4 Wild type 11.2 0.6 220 Brd4 Y97A 18.2 1.1 200 Brd4 Y390A 19.2 1.4 180 1e-1 1 1e1 1e2 1e3 Brd4 Double Mut > 1000 [Brd4] (nM) � Titrations of Brd4, Brd4 Y97A, Brd4 Y390A and Brd4 Y97A Y390A against 5nM AF647 benzodiazepine ligand � n=6 (replicates) 5
I-BET151 to BD1 Selective I-BET %I Conc. Brd4 BD1 Brd4 BD2 IC 50 nM (n) 32 (4) 6300 (4) � 200-fold domain selectivity demonstrated within this series � Probe quality and utility is highly dependent on selectivity Selective I-BET � Platform approach to screening assays has enabled a wide range of profiles to be identified and optimised � DiscoveRx BROMOscan panel confirms target class wide selectivity Increasing BET selectivity 6
Wider Bromodomain Target Class � >50 bromodomains � In isolation or combination with other domains � Multiple “untapped” opportunities for clinical utility � Rapid entry into Probes to confirm phenotype Structural divergence outside of the AcK pocket ZA Loop BC Loop 7
Generation of a Hit-ID platform for Bromodomains N N N O N N O H � Knowledge of key ligand-protein interactions derived from the BET programme compounds � Generation of a pharmacophore model � Selection of a focused screening set N O N H O N N N O O � Confirmation of the binding mode � Creation of bespoke fragment set that binds in the using crystallography AcK recognition pocket of the bromodomain Extending Fragment Based Discovery >40 Fragments crystallised 1400 Fragments screened Key Structural waters identified Pharmacophore refined Chung et al - Journal of Medicinal Chemistry (2012), 55(2), 576-586 8
Utilising Structural Knowledge to Develop Assays � Small molecule inhibitor crystallised with PCAF bromodomain � Derivative tagged with AlexaFluor488 Derivative tagged with AlexaFluor488 � FP ligand binding shown to be competed with acetylated peptides, acetyl lysine � Kd for PCAF BRD mutant ~ 100-fold higher than for WT 280 260 240 240 PCAF Brd 220 Mutant mP 200 PCAF Wild 180 Type 160 140 120 0.01 0.1 1 10 100 [PCAF] (uM) Expansion into Bromodomains with close homology 380 360 PCAF 340 320 320 BPTF BPTF 300 ZMY11 mP 280 260 CECR2 240 GCN5 220 200 ATAD2A 180 160 1e-4 1e-3 1e-2 1e-1 1 1e1 [Novel Brd] (uM) � FP tagged ligand profiled against Bromodomain containing proteins with close domain homology � Additional utility demonstrated for GCN5, FALZ (BPTF) and CECR2 9
Development & Utility of a Ligand Based Library � Utilised GSK’s Bromodomain chemical and structural knowledge to build an extensive library of small molecule ligands � Profiled ligands against broad panel of Bromodomain containing proteins Profiled ligands against broad panel of Bromodomain containing proteins � Hits for multiple targets; cross reactivity gives broad target class coverage � Screening assays developed and novel hits identified � Probe discovery platform capability established � Strategic collaboration initiated with DiscoveRx Inc. Epinova Chemical Probe Strategy Target Target Selectivity S l ti it Engagement Target Selection & Probe ID Target Target Effect Phenotype 10
Summary � Phenotypic screening led to the identification of a novel druggable epigenetic target class � Further refinements have delivered potent, selective BET inhibitors � Identifying novel areas of biology for drug discovery programmes requires the identification of high quality chemical probes � T Target class knowledge has enabled the establishment of a high t l k l d h bl d th t bli h t f hi h quality Bromodomain inhibitor identification platform � Breadth of knowledge and approaches has enabled rapid and efficient identification of novel potent and selective probes across the bromodomain family Acknowledgments � Rab Prinjha & colleagues within the Epinova DPU � Stuart Baddeley , Chun-wa Chung , Inmaculada Rioja, Chris Wellaway, Peter Craggs, Ryan Bingham, Laurie Gordon, Natalie Barrett, Paul Bamborough, Lee Harrison, David Hirst and Hannah Traynor � Many colleagues within GSK across MDR Chemical and Biological Sciences � DiscoveRx Inc – Paul Gallant, Daniel Treiber Kevin Lee 11
12
Recommend
More recommend
