Discovery of Epigenetic Chemical Probes Peter J. Brown Project Manager, Epigenetics Structural Genomics Consortium Sept 29...
Outline <ul><li>What is a chemical probe? </li></ul><ul><li>What are the goals of the Epigenetics Project? </li></ul><ul><...
What is a chemical probe? <ul><li>Chemical probes are potent, selective ligands which are used to validate a target using ...
Histone DNA Lysine Modification Write Read Erase Acetyl (Ac) HAT Bromo HDAC Methyl (Me n ) HMT Royal KDM Chemical Biology ...
Structural Genomics Consortium Epigenetics Project <ul><li>Global SGC Goal—to deliver 37 chemical probes in 4 years and pl...
Epigenetics Partners <ul><li>Pharma partners </li></ul><ul><ul><li>GlaxoSmithKline (GSK) </li></ul></ul><ul><ul><ul><li>Me...
Discovery Strategies <ul><li>High Throughput Screening </li></ul><ul><ul><li>Little overlap (670K unique compounds) </li><...
Discovery Strategies <ul><li>Fragment Screening </li></ul><ul><ul><li>Of all the target classes tested using this approach...
Discovery Strategies <ul><li>Structure-based Design </li></ul><ul><ul><li>Low/medium throughput screening of designed comp...
HMT Assays in production Struc solved by SGC Struc solved by others Assay in Production Assay in Development BIX-01294 B2 ...
HAT Assays in production Struc solved by SGC Struc solved by others Assay in Production B1 B2 B3 B4 B5 B6
Assessing Activity for HMTs Evys Collazo & Raymond C. Trievel et al. (2005) Analytical Biochemistry 342: 86–92
Protein Lysine Methyltransferase G9a <ul><li>First identified as a H3K9 MT in 2002  (Tachibana et al,  Genes Dev.  2002 , ...
Structure-based Design of UNC0224  BIX01294 G9a (ThioGlo): IC 50  = 180 nM G9a (AlphaScreen): IC 50  = 250 nM Kubicek, et ...
First Co-crystal Structure of G9a + small molecule: G9a-UNC0224 complex PDB code: 3K5K <ul><li>7-Dimethylaminopropoxy side...
Assay Measures Cellular Levels of H3K9me2 Dalia Barsyte (SGC), unpublished results <ul><li>In Cell Western assay in MDA-MB...
Compounds Designed  to Improve Cellular Potency <ul><li>Exploit newly identified SAR to increase lipophilicity, thus cell ...
UNC0638 More Potent Than BIX01294  Dalia Barsyte (SGC), unpublished results UNC0638 at 250 & 500 nM reduced cellular  H3K9...
Quantitative MS Analysis of  Effects of UNC0638 on Histone PTMs  Ben Garcia (Princeton) unpublished results <ul><li>UNC063...
UNC0638 Less Toxic Than BIX01294  Dalia Barsyte (SGC), unpublished results   In Vitro G9a IC 50  (nM) H3K9me2 48h IC 50  (...
Effects of UNC0638 in Other Cell Lines  Dalia Barsyte (SGC), unpublished results Cell Lines IC 50  (nM) Toxic/Func Ratio H...
UNC0638 Mechanism of Action  Tim Wigle, unpublished results UNC0638 is competitive with peptide, not SAM
Cocrystal Structure Confirms  MOA of UNC0638  <ul><li>UNC0638 occupies histone binding groove and does not interact with S...
UNC0638 Selectivity <ul><li>Epigenetic targets </li></ul><ul><ul><li>G9a:  K i  = 3 nM (Caliper), IC 50  < 15 nM (ThioGlo)...
UNC-SGC Probe Consortium Released UNC0638 as a Chemical Probe on June 1st  http://www.thesgc.org/chemical_probes/UNC0638/#...
Bromodomains <ul><li>Small domain (~110 residues) that selectively binds to acetylated lysine residues </li></ul><ul><li>B...
Bromodomain Goals <ul><li>At least 1 probe from each major subfamily </li></ul><ul><li><100 nM Kd by ITC or displacement a...
Tm shift data for Bromodomains <ul><li>Rapid, low protein consumption </li></ul><ul><li>Partially validated by comparison ...
Temperature A Potent BRD4 Ligand  Tm=10˚C <ul><li>Inspired by the publication of a BRD  patent by Mitsubishi </li></ul>Ja...
Interaction Confirmed by ITC
Selectivity for the BET subfamily Thermal Shift Data
Crystallography <ul><li>Co-crystal structures solved at high resolution with BRD4 (1),  </li></ul><ul><li>BRD3 (1 & 2) and...
Cell-based Assay James Bradner NUT-BRD4 GFP NUT-GFP
BETsOFF inhibits Proliferation James Bradner
BETsOFF inhibits Tumour Growth James Bradner Day 1 Day 5 Day 11 Day 18
Probe Consortium Released JQ1/SGCBD01 as a Chemical Probe on Sept 24th  <ul><li>Advance publication available through Natu...
Summary <ul><li>Outlined strategy for the discovery of Epigenetic Chemical Probes. </li></ul><ul><li>Shown how collaborato...
Acknowledgements <ul><li>Toronto </li></ul><ul><li>Cheryl Arrowsmith </li></ul><ul><li>Masoud Vedadi </li></ul><ul><li>Mat...
Upcoming SlideShare
Loading in …5
×

Probes 2010

1,993 views

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
1,993
On SlideShare
0
From Embeds
0
Number of Embeds
26
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Probes 2010

  1. 1. Discovery of Epigenetic Chemical Probes Peter J. Brown Project Manager, Epigenetics Structural Genomics Consortium Sept 29 th 2010 SGC Oxford SGC Toronto SGC Stockholm
  2. 2. Outline <ul><li>What is a chemical probe? </li></ul><ul><li>What are the goals of the Epigenetics Project? </li></ul><ul><li>How are we going to find chemical probes? </li></ul><ul><li>What is the progress to date? </li></ul>
  3. 3. What is a chemical probe? <ul><li>Chemical probes are potent, selective ligands which are used to validate a target using in vivo or cell-based experiments. </li></ul><ul><li>Probes do not have to contain novel chemical templates or meet rigorous criteria for clinical use in humans. </li></ul><ul><li>However, probes need to be sufficiently bioavailable and/or cell penetrable for experiments to be meaningful. </li></ul>
  4. 4. Histone DNA Lysine Modification Write Read Erase Acetyl (Ac) HAT Bromo HDAC Methyl (Me n ) HMT Royal KDM Chemical Biology of Epigenetics
  5. 5. Structural Genomics Consortium Epigenetics Project <ul><li>Global SGC Goal—to deliver 37 chemical probes in 4 years and place these in public domain for use in target validation experiments. </li></ul><ul><ul><li>Fully characterize probes in terms of </li></ul></ul><ul><ul><ul><li>Potency (IC 50 <100 nM) </li></ul></ul></ul><ul><ul><ul><li>selectivity across protein family (30-fold inter-branch selectivity) </li></ul></ul></ul><ul><ul><ul><li>activity in secondary and/or cell-based assays (<1  M) </li></ul></ul></ul><ul><ul><ul><li>determine structure of ligand-protein complex </li></ul></ul></ul><ul><li>Toronto group primarily focused on HMTs and HATs with some effort in the Royal family (12 probes). </li></ul><ul><li>Oxford group primarily focused on Demethylases and Bromo domains (25 probes). </li></ul>
  6. 6. Epigenetics Partners <ul><li>Pharma partners </li></ul><ul><ul><li>GlaxoSmithKline (GSK) </li></ul></ul><ul><ul><ul><li>Medicinal chemistry and Encoded Library Technology (ELT). </li></ul></ul></ul><ul><ul><li>Pfizer </li></ul></ul><ul><ul><ul><li>Medicinal Chemistry and high throughput screening. </li></ul></ul></ul><ul><ul><li>Novartis </li></ul></ul><ul><ul><ul><li>Medicinal Chemistry and cell-based assay development. </li></ul></ul></ul><ul><li>Academic/Government </li></ul><ul><ul><li>Center for Integrative Chemical Biology and Drug Discovery, University of North Carolina (UNC) </li></ul></ul><ul><ul><ul><li>Screening and medicinal chemistry </li></ul></ul></ul><ul><ul><li>NIH Chemical Genomics Center (NCGC) </li></ul></ul><ul><ul><ul><li>High throughput screening and chemistry </li></ul></ul></ul><ul><ul><li>Ontario Institute for Cancer Research (OICR) </li></ul></ul><ul><ul><ul><li>Medicinal chemistry </li></ul></ul></ul>
  7. 7. Discovery Strategies <ul><li>High Throughput Screening </li></ul><ul><ul><li>Little overlap (670K unique compounds) </li></ul></ul><ul><ul><li>Multiple screening formats available </li></ul></ul>NCGC 440K UNC 110K OICR 150K
  8. 8. Discovery Strategies <ul><li>Fragment Screening </li></ul><ul><ul><li>Of all the target classes tested using this approach, the highest success rate is for enzyme inhibitors. </li></ul></ul><ul><ul><li>Screening of 2040 low MW fragments at high concentration. </li></ul></ul><ul><ul><li>84% of collection MW between 150 and 250. </li></ul></ul><ul><ul><li>Screen at 300  M in activity assay. </li></ul></ul><ul><ul><li>X-ray follow-up. </li></ul></ul>
  9. 9. Discovery Strategies <ul><li>Structure-based Design </li></ul><ul><ul><li>Low/medium throughput screening of designed compounds or small arrays </li></ul></ul><ul><li>Virtual Screening </li></ul><ul><ul><li>Assessment of 6M diverse compounds using MolSoft software suite. </li></ul></ul><ul><ul><li>Additional virtual arrays built around HTS hits . </li></ul></ul>
  10. 10. HMT Assays in production Struc solved by SGC Struc solved by others Assay in Production Assay in Development BIX-01294 B2 B3 B4 B5 B6 B7 B1
  11. 11. HAT Assays in production Struc solved by SGC Struc solved by others Assay in Production B1 B2 B3 B4 B5 B6
  12. 12. Assessing Activity for HMTs Evys Collazo & Raymond C. Trievel et al. (2005) Analytical Biochemistry 342: 86–92
  13. 13. Protein Lysine Methyltransferase G9a <ul><li>First identified as a H3K9 MT in 2002 (Tachibana et al, Genes Dev. 2002 , 16, 1779) . Shares 80% sequence identity with GLP (AKA EHMT1) (Chang et al, Nat. Struct. Mol. Biol. 2009, 16, 31). G9a and GLP work cooperatively via forming a heterodimer. </li></ul><ul><li>Overexpressed in cancer and knockdown inhibits cancer cell growth (McGarvey et al, Cancer Res. 2006, 66, 3541) (Kondo et al, PLoS ONE 2008, 3, e2037) </li></ul><ul><li>Also dimethylates K373 of p53, which results in the inactivation of p53 (Huang et al, J. Biol. Chem. 2010 , 285, 9636) </li></ul><ul><li>Plays an important role in the development of cocaine addition (Maze et al, Science 2010, 327, 213) and mental retardation (Schaefer et al, Neuron 2009, 64, 678), and in maintenance of HIV-1 latency (Imai et al, J. Biol. Chem. 2010 , 285, 16538) </li></ul><ul><li>BIX01294, a G9a inhibitor, efficacious as a replacement for c-Myc and Sox2 for reprogramming of mouse fetal neural precursor cells into iPS cells (Shi et al, Cell Stem Cell 2008, 3, 568) </li></ul>
  14. 14. Structure-based Design of UNC0224 BIX01294 G9a (ThioGlo): IC 50 = 180 nM G9a (AlphaScreen): IC 50 = 250 nM Kubicek, et al. 2007, Mol Cell, 473 G9a (ThioGlo): IC 50 = 43 nM G9a (AlphaScreen): IC 50 = 57 nM G9a (ITC): K D = 23 nM UNC0123 G9a (ThioGlo): IC 50 = 330 nM G9a (AlphaScreen): IC 50 = 230 nM Reduced MW while maintaining potency Array-based Optimization* UNC0224 GLP-BIX01294 complex Adopted from Chang, et al. 2009, Nat. Stru. Mol. Bio., (16), 316 Liu et al, J. Med. Chem. 2009 , 52, 7950 Synthesized in 10 steps
  15. 15. First Co-crystal Structure of G9a + small molecule: G9a-UNC0224 complex PDB code: 3K5K <ul><li>7-Dimethylaminopropoxy side chain binds in the lysine binding channel, validating the binding hypothesis, but does not fully fill available space </li></ul>Liu et al, J. Med. Chem. 2009 , 52, 7950
  16. 16. Assay Measures Cellular Levels of H3K9me2 Dalia Barsyte (SGC), unpublished results <ul><li>In Cell Western assay in MDA-MB-231 cells </li></ul><ul><ul><li>Treated with inhibitors for 48 h </li></ul></ul><ul><ul><li>Cellular H3K9me2 levels measured by immunostaining using anti-H3K9me2 & IR800 </li></ul></ul><ul><li>Simultaneously assess cell viability via DNA staining using DRAQ5 </li></ul><ul><li>Although UNC0224 is more potent than BIX01294 in biochemical assays, it is less potent in the cell-based assay </li></ul>
  17. 17. Compounds Designed to Improve Cellular Potency <ul><li>Exploit newly identified SAR to increase lipophilicity, thus cell membrane permeability while maintaining high potency </li></ul><ul><li>Prepared > 50 combination compounds. Aiming to achieve balanced in vitro potency & physical chemical propertie s </li></ul>Feng Liu & Xin Chen
  18. 18. UNC0638 More Potent Than BIX01294 Dalia Barsyte (SGC), unpublished results UNC0638 at 250 & 500 nM reduced cellular H3K9Me2 levels close to G9a/GLP knockdown
  19. 19. Quantitative MS Analysis of Effects of UNC0638 on Histone PTMs Ben Garcia (Princeton) unpublished results <ul><li>UNC0638 ↓H3K9me2, –H3K9me3, ↓H3K9me1, ↑H3K9un & ↑H3K14ac </li></ul><ul><li>Similar to G9a and GLP Knockdown </li></ul>520 521 522 523 524 525 526 527 528 m/z 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Relative Abundance 521.307 521.808 523.822 524.323 526.337 522.309 526.839 524.824 522.810 527.340 Control shRNA shRNA G9a shRNA G9a+GLP1 H3K9me2 528 529 530 531 532 533 534 535 m/z 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Relative Abundance 530.829 528.314 533.345 531.331 528.815 533.846 531.832 529.317 534.347 Control shRNA shRNA G9a shRNA G9a+GLP1 H3K9me3 520.5 521.0 521.5 522.0 522.5 523.0 523.5 524.0 524.5 525.0 525.5 526.0 526.5 527.0 527.5 528.0 m/z 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Relative Abundance 521.306 521.807 523.821 524.323 526.337 522.308 526.838 524.824 522.810 527.339 Control BIX UNC 638 H3K9me2 528.0 528.5 529.0 529.5 530.0 530.5 531.0 531.5 532.0 532.5 533.0 533.5 534.0 534.5 535.0 m/z 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Relative Abundance 528.314 530.829 533.345 528.815 531.331 533.846 529.317 531.832 534.347 Control BIX UNC 638 H3K9me3
  20. 20. UNC0638 Less Toxic Than BIX01294 Dalia Barsyte (SGC), unpublished results   In Vitro G9a IC 50 (nM) H3K9me2 48h IC 50 (nM) Cell Tox 48h EC 50 (nM) Tox/Func Ratio BIX-01294 133 ± 15 500 ± 43 2805 5.6 UNC0638 <15 81 ± 9 11190 138 Poor separation of functional and toxic effects BIX01294 10 0 10 1 10 2 10 3 10 4 10 5 0 10 20 30 40 50 60 70 80 90 100 110 H3K9m2 MTT nM % response Good separation of functional and toxic effects UNC0638 H3K9m2 MTT
  21. 21. Effects of UNC0638 in Other Cell Lines Dalia Barsyte (SGC), unpublished results Cell Lines IC 50 (nM) Toxic/Func Ratio H3K9me2 MTT Breast carcinoma MDA231 81 11,000 138 MCF7 70 7,600 109 Prostate carcinoma PC3 59 14,000 233 22RV1 48 4,500 95 Colon carcinoma HCT116 wt 210 11,000 55 HCT116 p53-/- 240 11,000 47 Human fibroblast IMR90 120 2,300 19
  22. 22. UNC0638 Mechanism of Action Tim Wigle, unpublished results UNC0638 is competitive with peptide, not SAM
  23. 23. Cocrystal Structure Confirms MOA of UNC0638 <ul><li>UNC0638 occupies histone binding groove and does not interact with SAM binding pocket. Same binding mode as BIX01294 & UNC0224 </li></ul>UNC0224 (PDB 3K5K) BIX-01294 (PDB 3FPD) UNC0638 (PDB 3NNI) PDB Code: 3NNI Masoud Vedadi (SGC) UNC0638 SAH
  24. 24. UNC0638 Selectivity <ul><li>Epigenetic targets </li></ul><ul><ul><li>G9a: K i = 3 nM (Caliper), IC 50 < 15 nM (ThioGlo); GLP: IC 50 = 19 nM (ThioGlo) </li></ul></ul><ul><ul><li>Other PKMTs: inactive vs SETD7, MLL, SMYD3, SUV39H1, SUV39H2, EZH2, DOT1L, SETD8, PRDM1, PRDM10, PRDM12 </li></ul></ul><ul><ul><li>PRMTs: inactive vs PRMT1, PRMT3 </li></ul></ul><ul><ul><li>PKDMs: JMJD2E, IC 50 = 4660 nM </li></ul></ul><ul><ul><li>HATs: inactive vs HTATIP </li></ul></ul><ul><ul><li>DNMTs: DNMT1, IC 50 = 1287 nM (to be confirmed) </li></ul></ul><ul><li>Non-epigenetic targets: a panel of 29 receptors, ion channels & transporters </li></ul><ul><ul><li>< 30% inhibition at 1  M versus 26 targets </li></ul></ul><ul><ul><li>> 30% inhibition at 1  M versus 3 targets: M 2 (64%),  1A (90%),  1B (69%) </li></ul></ul><ul><li>A broad panel of 40 GPCRs </li></ul><ul><ul><li>< 50% inhibition at 1  M versus 34 targets </li></ul></ul><ul><ul><li>> 50% inhibition at 1  M versus 6 targets: 5HT 2A (75%),  1A (55%),  2C (65%), M 1 (80%), M 2 (99%), & M 4 (87%) </li></ul></ul><ul><ul><li>K i (nM) = 1,300 (5HT 2A ), 730 (  1A ), 290 (  2C ), 960 (M 1 ), 1,400 (M 2 ), & 1,100 (M 4 ) </li></ul></ul><ul><ul><li>At least ~100-fold selective for G9a over GPCRs tested </li></ul></ul><ul><li>Inactive versus 24 kinases </li></ul>
  25. 25. UNC-SGC Probe Consortium Released UNC0638 as a Chemical Probe on June 1st http://www.thesgc.org/chemical_probes/UNC0638/#overview <ul><li>Data released prior to publication. Living document that is updated as new data is generated. </li></ul><ul><li>UNC0638 is available through Sigma-Aldrich </li></ul>
  26. 26. Bromodomains <ul><li>Small domain (~110 residues) that selectively binds to acetylated lysine residues </li></ul><ul><li>Bundle of four  -helices, Z, A, B and C plus two loops forming a pocket with a conserved Asn residue </li></ul><ul><li>A recognition domain forming part of numerous chromatin modifying proteins, including histone acetyltransferases (eg CREB & PCAF) and transcriptional coactivators/repressors </li></ul><ul><li>57 unique bromodomains identified to date </li></ul>
  27. 27. Bromodomain Goals <ul><li>At least 1 probe from each major subfamily </li></ul><ul><li><100 nM Kd by ITC or displacement assay </li></ul><ul><li>>30-fold selectivity vs representative proteins from other subfamilies (highlighted in red) </li></ul><ul><li>Demonstration of interaction with target protein in cells at <1uM </li></ul>
  28. 28. Tm shift data for Bromodomains <ul><li>Rapid, low protein consumption </li></ul><ul><li>Partially validated by comparison with ITC </li></ul><ul><li>Screened ~10K compounds: fragments, VLS, pharmacophores </li></ul><ul><li>Hits for >10 bromodomains </li></ul>Thermal stability (DSF) Bromo Subfamily 1 2 3 5 6 7 8 9 13 14 Bromodomain SMARCA2 LOC93349 BAZ2B ATAD2 BRD9 BRPF1 CREBBP BRD2 BRD4 CECR2 PCAF FALZ Most potent hit  Tm / ˚ C 5.6 12.9 3.0 3.8 5.6 12.1 6.7 4.8 8.4 3.6  5.3 3.9
  29. 29. Temperature A Potent BRD4 Ligand  Tm=10˚C <ul><li>Inspired by the publication of a BRD patent by Mitsubishi </li></ul>James Bradner
  30. 30. Interaction Confirmed by ITC
  31. 31. Selectivity for the BET subfamily Thermal Shift Data
  32. 32. Crystallography <ul><li>Co-crystal structures solved at high resolution with BRD4 (1), </li></ul><ul><li>BRD3 (1 & 2) and BRD2 (1) </li></ul><ul><li>Excellent shape complementarity </li></ul>
  33. 33. Cell-based Assay James Bradner NUT-BRD4 GFP NUT-GFP
  34. 34. BETsOFF inhibits Proliferation James Bradner
  35. 35. BETsOFF inhibits Tumour Growth James Bradner Day 1 Day 5 Day 11 Day 18
  36. 36. Probe Consortium Released JQ1/SGCBD01 as a Chemical Probe on Sept 24th <ul><li>Advance publication available through Nature </li></ul><ul><li>JQ1/SGCBD01 is available from the lab of Dr. James Bradner </li></ul>http://www.thesgc.org/chemical_probes/JQ1_SGCBD01/#overview
  37. 37. Summary <ul><li>Outlined strategy for the discovery of Epigenetic Chemical Probes. </li></ul><ul><li>Shown how collaborators are key components in the project. </li></ul><ul><li>Significant progress made in identifying multiple chemical starting points for optimization </li></ul><ul><li>Two probes declared for G9a/GLP and BRD BET subfamily. </li></ul>
  38. 38. Acknowledgements <ul><li>Toronto </li></ul><ul><li>Cheryl Arrowsmith </li></ul><ul><li>Masoud Vedadi </li></ul><ul><li>Matthieu Schapira </li></ul><ul><li>Jinrong Min </li></ul>Oxford Chas Bountra Tom Heightman Stefan Knapp Brian Marsden F UNDING P ARTNERS Canadian Institutes for Health Research, Canadian Foundation for Innovation, Genome Canada through the Ontario Genomics Institute, GlaxoSmithKline, Knut and Alice Wallenberg Foundation, Merck & Co., Inc., Novartis Research Foundation, Ontario Innovation Trust, Ontario Ministry for Research and Innovation, Swedish Agency for Innovation Systems, Swedish Foundation for Strategic Research, and Wellcome Trust. Stephen Frye Bill Janzen Jian Jin Bryan Roth Tim Willson Ryan Trump Anton Simeonov James Bradner Jun Qi Ben Garcia

×