Word - Babraham_Institute - Discovery Biology for Biomedicine
Fragments for drug discovery and chemical biology
-
Upload
warwickamr -
Category
Science
-
view
259 -
download
3
Transcript of Fragments for drug discovery and chemical biology
1
Fragments for drug discovery and chemical biology
Rod Hubbard
YSBL, York,
Vernalis, Cambridge
Warwick, Sep 2016
For slides – [email protected]
2
Finding small molecule hits
• Trying to find compounds that bind to target
• Compounds need to have required shape and chemistry
• High Throughput Screening
• Compounds decorated in the wrong way
• Particularly a problem with new target classes
Target
Hit
3
• Hits from fragments
• Find small parts that bind – usually biophysical screen
• Then grow or merge fragments to create hit compound
• Usually structure-guided design
Why fragments?
Hit
Target
Screen Structure
guided
design
4
Why fragments?
• Hits from fragments
• Find small parts that bind
• Then grow or merge fragments to create hit compound
• Can also generate ideas
• Deconstruct other hits to optimise key interaction motifs
• Suggest interactions to exploit in hit / lead optimisation
• Scaffold hopping
Hit
Target
5
Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
Warwick, Sep 2016
6
Why are fragments different?
• A fragment is just a small weak hit
• Requires assay(s) that can detect binding reliably
• Methods for evolving fragments (libraries and/or design)
• Design of library includes constraints of assay / evolution
Affinity
10mM 1mM 100mM 10mM 1mM
Fragments MW 110-250
Scaffolds MW 250-350
Lead Compounds
Hit Compound MW 250-500
Warwick, Sep 2016
7
Screening fragment libraries
• Different experimental approaches have different strengths and limitations
Warwick, Sep 2016
Affinity 10mM 1mM 100mM 10mM 1mM
Fragments MW 110-250
Scaffolds MW 250-350
Lead Compounds
X-Ray crystallography
Ligand-observed NMR
Surface Plasmon Resonance (SPR)
Enzyme / binding assays (HCS)
Isothermal Titration Calorimetry (ITC)
Hit Compound MW 250-500
Protein-observed NMR
Differential scanning fluorimetry (DSF / TSA)
Hubbard & Murray (2011), Meth Enzym, 493: 509; Meiby et al (2013) Anal Chem 85: 6756
Mass spectrometry (MS)
Weak Affinity Chromatography (WAC)
8
Optimise fragment Fragment to hit : SAR by catalog off-rate screening
-5
0
5
10
15
20
25
30
-50 0 50 100 150 200 250 300
RU
Resp
on
se
Tim e s
Cycle: 95 VER-00082099i 50 nM
Fitted Cycle: 95 VER-00082099i
Cycle: 96 VER-00082099i 500 n
Fitted Cycle: 96 VER-00082099i
Cycle: 97 VER-00082099i 5000
Fitted Cycle: 97 VER-00082099i -4
-2
0
2
4
6
8
-100 -50 0 50 100 150 200 250 300
RU
Resp
on
se
Tim e s
Cycle: 103 VER-00055030l 50 n
Fitted Cycle: 103 VER-00055030l
Cycle: 104 VER-00055030l 500
Fitted Cycle: 104 VER-00055030l
Cycle: 105 VER-00055030l 5000
Fitted Cycle: 105 VER-00055030l
Characterisation
X-ray or NMR guided model
The process
Warwick, Sep 2016
Hubbard et al (2007), Curr Topics Med Chem, 7: 1568 Hubbard and Murray (2011), Methods Enzym, 493: 509
Target
Optimise fragment
Hits
Competitive NMR screen Fragment Library
~ 1500 compounds Ave MW 190
Design, Build & Test
N
N SNH2
O
NH
Cl
Cl
ON
NO
OH
OH
O
NH
N O
NN
NH2
OOMe
NN
NH2
SNH
O
N
N
N
NH2
Cl
ClN
SNNH2
NH2
N
NH2
O
OEt
Virtual screen; literature; library screen Screen by SPR, DSF, WAC,
biochem assay, Xray
Drug?
9
Finding fragments
• Finding fragments that bind is not difficult • A good way of assessing target “ligandability”
• Low hit rate can indicate difficult to progress • See also Hajduk (2005) J Med Chem, 48, 2518
• Low hit rates from catenin, Pin1 and Hsp70 ATP site
Updated from Chen & Hubbard (2009), JCAMD, 23: 603
0%
1%
2%
3%
4%
5%
6%
7%
8%
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Dscore
Cla
ss 1
hit
s ra
tes
Low hit rates (< 2%) High hit rates (> 2%)
Kinases
high hit rate
1 Calculate druggability
Val
idat
ed h
it r
ate
Ligandability calculated from structure (DScore)
Validated
hit rate
from a
ligand-
observed
NMR
screen
protein-protein interaction
targets – varying hit rates
Poor targets
10
Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
Warwick, Sep 2016
11
Pre-Clinical Clinical Trials
Drug Discovery
Warwick, Sep 2016
Discovery
I II III
Patient
Target Hit ID
H2Ls
Lead
Optimisation
Good Idea
Hit Compound
Lead Compound
Clinical Candidate
Drug
Medicine
12
Hsp90 – the Target
• Hsp90 function is complex involving protein dimerisation and
association to other chaperones and co-factors
GM Geldanamycin – a fungal natural product
Hsp90 requires ATP to function (via hydrolysis to ADP)
Kamal et al (2004) Trends in Mol Med 10: 283
13
Hsp90: Fragment screen
• Targetting the N-terminal domain – an ATPase
• FBLD programme began in early 2002 • - screened library of 729 fragments by NMR
• 17 fragments identified • Crystal structures for most fragments binding to Hsp90
Amide Amino-pyrimidine Second site binder
Resorcinol
14
Using fragments
• Finding fragments that bind is not difficult
• The challenge is knowing what to do with the hits • Link, grow or merge
Screen GROW
Warwick, Sep 2016
15
Hsp90 – AUY922 story
Starting fragment Hit from SAR by
Catalogue (also MTS and VS)
• GI50 in HCT116 colon cell line
FP IC50 = 0.28mM
GI50 = 6mM
FP IC50 = 0.009mM
GI50 = 0.014mM
FP IC50 = ~1mM
D93 G97
K58
F138
L107
rCat N
O
O
O
O
Phase II Candidate
(Novartis)
D93
G97
K58
F138
L107
Brough et al (2008) J Med Chem 51,196-218 Roughley et al (2012) Top Curr Chem , 317, 61
16
Using fragments
• Finding fragments that bind is not difficult
• The challenge is knowing what to do with the hits • Link, grow or merge
Known Ligands
Virtual Screening hits Screen
Detailed Design
17
NN
NH2
OOMe
VER-26734
FP IC50>5mM
NN
NH2
SNH
OVER-52959
FP IC50=535mM
Fragment Evolved fragment
HSP90 – BEP800 story
N
N
N
NH2
Cl
ClN
NO
OH
OH
O
NH
N O
luminespib
Vernalis Phase II candidate (FBLD
/ SBDD derived)
VER-41113
FP IC50=1.56mM
Virtual Screening Hit
VER-45616
FP IC50=0.9mM
SNNH2
NH2
N
NH2
O
OEt
Virtual Screening Hit
Brough et al (2009) J Med Chem 52,4794-4809 Roughley et al (2012) Top Curr Chem 317, 61
D93 G97 K58
F138 L107
18
N
N SNH2
O
NH
Cl
Cl
ON
NN
NH2
OOMe
VER-26734
FP IC50>5mM
NN
NH2
SNH
OVER-52959
FP IC50=535mM
Fragment Evolved fragment
HSP90 – BEP800 story
N
N
N
NH2
Cl
ClN
VER-82576
NVP-BEP800
FP IC50=0.058mM
KD = 0.9nM (SPR)
HCT116 GI50=0.161mM
BT474 GI50=0.057mM
NO
OH
OH
O
NH
N O
luminespib
Vernalis Phase II candidate (FBLD
/ SBDD derived)
VER-41113
FP IC50=1.56mM
Virtual Screening Hit
VER-45616
FP IC50=0.9mM
SNNH2
NH2
N
NH2
O
OEt
Virtual Screening Hit
Brough et al (2009) J Med Chem 52,4794-4809 Roughley et al (2012) Top Curr Chem 317, 61
19
Drug leads from fragments
• Vernalis have disclosed leads for: • kinases such as Chk1, DYRK1A, PDPK1, Pim1, Pak1, Pak4,
STK33 and PDHK
• ATPases such as DNA gyrase, HSP70 and HSP90
• protein-protein interaction targets Pin1, Mcl-1 and Bcl-2
• FAAH and tankyrase
• In recent years – clinical candidates for: • Chk1, FAAH, Hsp90, (Bcl-2 – inspired by), Mcl-1
• Many other examples published • Small and large pharma
• List on http://practicalfragmentsblogspotcom/.
• Some recent reviews and books
Warwick, Sep 2016
20
Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
Warwick, Sep 2016
21
Pre-Clinical Clinical Trials
Drug Discovery
Warwick, Sep 2016
Discovery
I II III
Patient
Target Hit ID
H2Ls
Lead
Optimisation
Good Idea
Hit Compound
Lead Compound
Clinical Candidate
Drug
Medicine
22
Pre-Clinical Clinical Trials
Chemical Biology
Warwick, Sep 2016
Discovery
I II III
Patient
Target Hit ID
H2Ls
Lead
Optimisation
Good Idea
Hit Compound
(Lead Compound)
Clinical Candidate
Drug
Medicine
23
Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
Warwick, Sep 2016
24
Warwick, Sep 2016 The TolB project
• A protein involved in bacterial attack / defence
• Structures determined of TolB in complex with various proteins (collaboration with Colin Kleanthous, now in Oxford)
• Exact mechanism of action still unclear
• Crystal structure with peptide bound
• Fragment screen (thermal shift) identified fragments that bind to the tryptophan pocket
• But also new site
25
Warwick, Sep 2016 The TolB project
• A protein involved in bacterial attack / defence
• Structures determined of TolB in complex with various proteins (collaboration with Colin Kleanthous, now in Oxford)
• Exact mechanism of action still unclear
• Crystal structure with peptide bound
• Fragment screen (thermal shift) identified fragments that bind to the tryptophan pocket
• But also new site
Sequence is conserved in this region
(not noticed before)
Mutation affects activity
26
Warwick, Sep 2016 Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
•Warwick, Sep 2016
27
Warwick, Sep 2016 Bacterial orthologue – BtGH84
• Family 84 glycoside hydrolase
• 40% identity with human OGA domain
• Crystal structure available
• Low micromolar (class-wide unselective)
inhibitors developed previously
• tool compounds
• Fragment screen of 91 fragments
• Ligand observed NMR
• PUGNAc as competitor
• Investigated both competitive and non-competitive inhibitors
PUGNAc Thiamet G
Darby et al Angew Chemie (2014) 53:13419
28
Warwick, Sep 2016
• Non-competitive fragment M1E05 enhances BtGH84 activity
• Confirmed with alternative substrate, many controls
• Activation is concentration dependent
• Increases the Vmaxapp and reduces the Km
app - 2.5 mM to 1.3 mM
BtGH84 is activated by a fragment
M1E05
Darby et al Angew Chemie (2014) 53:13419
29
Warwick, Sep 2016 BtGH84 is activated by a fragment
• Non-competitive fragment M1E05 enhances BtGH84 activity
• Confirmed with alternative substrate, many controls
• Activation is concentration dependent
• Increases the Vmaxapp and reduces the Km
app - 2.5 mM to 1.3 mM
• Assay, NMR and ITC confirm that M1E05 affects PUGNAc binding
M1E05
Darby et al Angew Chemie (2014) 53:13419
30
Warwick, Sep 2016 BtGH84 is activated by a fragment
• Non-competitive fragment M1E05 enhances BtGH84 activity
• Confirmed with alternative substrate, many controls
• Activation is concentration dependent
• Increases the Vmaxapp and reduces the Km
app - 2.5 mM to 1.3 mM
• Assay, NMR and ITC confirm that M1E05 affects PUGNAc binding
• Higher activity (200µM) compounds identified (see paper)
• Non-essential activator kinetics suggests mechanism
• Compounds stabilise folding of loop required for catalysis
• Better compounds required for probing GlcNAc status in cells
• But possibilities for industrial / bioprocessing enzymes
• Investigating covalent tethering of activators (it works !!)
M1E05
Darby et al Angew Chemie (2014) 53:13419
31
Warwick, Sep 2016 Overview
• Requirements for fragment-based discovery
• Examples of drug discovery • Hsp90
• Examples of chemical biology • Using fragments to explore binding
• Enzyme activators
• Probing the bacterial replisome
32
Warwick, Sep 2016 Fragnet
• EU ITN funded from March 2016
• 15 advanced PhD students employed across Europe to be trained in the methods of fragment based discovery
• Students in academic beneficiaries: VUA, Amsterdam, University of York, University of Barcelona, Hungarian Academy of Sciences
• Students in beneficiaries: Novartis, Switzerland, Vernalis, UK; Beactica, Sweden; Zobio, Holland
• Partners: GSK, UK, Iota, UK, Roche, Switzerland, Servier, France
• About 60% of the projects are developing methods
• Computational, novel synthesis, covalent targetting
• A couple of the applications are antimicrobial
• One of which is the bacterial replisome work at York
•
33
Warwick, Sep 2016 Bacterial replisome
• A collection of proteins / complexes that together replicate DNA
in bacteria
• Peter McGlynn at York able to reconstitute in vitro
34
Warwick, Sep 2016 Bacterial replisome
• A collection of proteins / complexes that together replicate DNA
in bacteria
• Peter McGlynn at York able to reconstitute in vitro
• Screen of 1000 member York library for inhibition of DNA
synthesis – a pseudo phenotypic screen
• Identified DNA intercalators through topoisomerase assay
• Now running functional assays to identify possible mechanisms
• Crystal structures available for many of the components / sub-
complexes
• Target identification underway