Drug Discovery Overview

Drug Discovery Overview
Hits to the Clinic
1 Chatterjee – Lecture at TSRI Oct 29 2014
Outline
•
Medicinal chemistry strategies
– Physiochemical properties – General terms
– On-target
g versus off-tissue effects
– Target-based optimization case study
• Mdm-2/p53
– Cell-based optimization case study
• Malaria
– Pharmacokinetic optimization
Leeson and
Sprinhthorpe
Nature Reviews Drug
Discovery, 2007, 881890.
• Effects on cardiotoxicity (S1P1 agonists)
•
Take home messages
– Principals for optimization
• Structure-based and physiochemical property based
• Pharmacokinetic based (and interplay with toxicity)
– Understanding compound progression schemes
• Lots of compounds so need clear benchmarks for success (attrition-based
strategy)
• Close interplay between chemistry, biology, pharmacology, toxicology
2 Chatterjee – Lecture at TSRI Oct 29 2014
S1P1 Receptor and Immune Diseases
Mechanism of Action
p agonists
g
such as FTY720 ((spingolipids)
p g p )
• S1P Receptor
leads to S1P1 down-regulation and prevents lymphocyte
egress. Clin Neuropharmacol. 2010 ; 33(2): 91–101.
Key issues
• On-target bradychardia (slow heart rate defined as less
than 60 bpm)
p ) upon
p first dosing
g
– Likely a Cmax effect
• Truly on-target toxicity? S1P3 receptor
• Can one modulate this effect my pushing out Tmax and
“flattening the PK curve”?
3 Chatterjee – Lecture at TSRI Oct 29 2014
Physicochemical properties for BBB permeability
Multiple-color band Radar chart, a tool for
lead optimization. Here the combined shaded
area represents the property ranges of 95% of
approved CNS drugs.
drugs The green band
represents the most densely populated
property range containing 50% of the
approved CNS drugs. The red line represents
the properties of the test molecule
(amitriptyline, a tertiary amine tricyclic
antidepressant, is structurally related to both
the skeletal muscle relaxant cyclobenzaprine
and the thioxanthene antipsychotics such as
thiothixene). During lead optimization, it is
often found that even after ROF type rules
have been satisfied the lead optimization
process does not achieve the goal. In such
cases, optimizing the physicochemical
properties
ti toward
t
d the
th green region
i may have
h
a better chance of success.
< ACS Chem.
Chem Neurosci.
Neurosci 2012,
2012 3
3, 50−68 >
4 Chatterjee – Lecture at TSRI Oct 29 2014
Check points for the CNS drugs to Penetrate Blood Brain Barrier
Free Drug Hypothesis
(total clearance adjusted for free fraction)
5 Chatterjee – Lecture at TSRI Oct 29 2014
Structure-based Design
•
p53/MDM-2 interaction blocker
– p53 is key tumor suppressor and 1st gen compound mimic MDM-2
peptide (though the interactions span ~120 residues and subsequent
peptides
tid d
do nott h
have good
d orall exposure))
– HTS led to nutlins and subsequent optimization
• Top. Med Chem 2012, 8; 57-80.
Flexible pocket
6 Chatterjee – Lecture at TSRI Oct 29 2014
MDM2/p53
•
10 compounds in clinical trials
–
–
–
–
–
–
Phase1/2- ALRN-6924 (peptide)
Phase 1 – RG-7775
Phase 1 – HDM-201/CGM-097
HDM 201/CGM 097 (Novartis)
Phase 1 – RO-6839921
Phase 1 – AMG-232
Phase 1 – SAR-405838
7 Chatterjee – Lecture at TSRI Oct 29 2014
Protease inhibitors
Examples of efficient bond-construction
HCV NS3/4 p
protease inhibitor ((MK-7009))
Convergent olefin
cross-metathesis
O
NH
N
8 Chatterjee – Lecture at TSRI Oct 29 2014
O
O
HN
NH
N
O
S1P1 binding sites
Science 2012, 335, 851
9 Chatterjee – Lecture at TSRI Oct 29 2014
Current Clinical Compounds
FTY720
BAF312 (Phase 3 – Novartis)
200 nM
Bradyarrhythmia
y
y
at 10 mg/day
g y
SEW2871 (HTS hit J. Biol. Chem 2004, 13839.)
WO 2011/060392
10 Chatterjee – Lecture at TSRI Oct 29 2014
WO 2012/158550
20 pM
20h human t1/2
Tmax pushed out
(1 7h versus 7h)
(1.7h
Cell-based Optimization of Novel Antiparasitics
11 Chatterjee – Lecture at TSRI Oct 29 2014
Drug Discovery Infrastructure
Target
Validation
HTS
Exploratory
Chemistry
Candidate
Selection Point
Full Lead
Optimization
Clinic
D0
D1
Lead Discovery
•
Large, diverse compound library
•
Fully robotic uHTS system capable
of >2 million data points/day
•
Automated compound profiling
system
•
Screening informatics infrastructure
(LDDB)
12 Chatterjee – Lecture at TSRI Oct 29 2014
D2

•
•
•
•
D3
Medicinal & Analytical Chemistry
Automated library synthesis
Experienced medicinal chemistry
staff
High-throughput analytical and
purification technology
Sophisticated compound
management
Tox studies

•
•
•
Pharmacology/ADMET
In vitro and in vivo ADMET
capabilities
State-of-the-art bioanalytical
technology
In vivo efficacy models in
metabolic disease, immunology
and oncology
Infectious Diseases of the Developing World
•
Objective
– Identification of new pharmacophore drug candidates
to improve treatment options for infectious diseases in
low and middle income countries
•
Scientific Rationale
– Prioritized disease targets where current drugs are
toxic/ ineffective and current drug discovery activities
are low relative to medical burden
– Deprioritized where other approaches are likely to be
more effective (vaccines, sanitation)
– Programs focused on Malaria
Malaria, Tuberculosis
Tuberculosis,
Kinetoplastid Diseases
•
Approach
– Reduction in burden of infectious agent will lead to
improved clinical outcomes
– Reliance on phenotypic screens for hit finding
– Deconvolution of mechanism of resistance through
genetic selection
13 Chatterjee – Lecture at TSRI Oct 29 2014
Mosquito
P.falciparum
Human
M.tuberculosis
Sand fly
L.donovani
Kissing bug
T.cruzi
Tsetse fly
T.brucei
Scope and Scientific Approaches
•
Whole pathogen cell-based screening approach
–
–
–
–
•
Majority of anti-infectives discovered through cell-based screening
• Facilitates simultaneous interrogation of multiple nodes/ pathways
• Consideration of compound permeability and metabolism in-built
Target-based approaches have been less successful, including internal programs
HTS on >2 million compounds for all pathogens, with broad compound profiling in secondary
assays
Closely coordinated chemistry, pharmacology and biology teams
Target identification methods
–
–
–
Lab-evolved
ab e o ed sstrains
a s resistant
es s a to
o co
compound
pou d treatment
ea e followed
o o ed by ge
genetic
e ca
analysis
a ys s
Promoter signature profiling
Compound affinity pull-downs
• Ensures no shared in vitro resistance mechanism with current drugs
• Target identification enables biochemical optimization approach for follow-on compounds
Presentation will highlight work in Malaria, Leishmania and Chagas Disease
14 Chatterjee – Lecture at TSRI Oct 29 2014
Malaria Medical Need
•
500
00 million
illi cases annually
ll worldwide
ld id
–
•
Resistance to current therapies widespread with exception of artemisinin derivatives
–
–
•
Nearly 1 million deaths/yr, >75% under age of 5
Increased parasite clearance times for Artemisinin-based therapies seen in Thai-Cambodian border
Current artemisinin-based combination therapies (ACTs) contraindicated in 1st trimester
No current therapies address need for blood-stage, liver-stage, and transmission
blocking activity
Re-infects
mosquitoes
Overview of the malaria
parasite life cycle
Only
primaquine
works
k on
hypnozoites
15 Chatterjee – Lecture at TSRI Oct 29 2014
The NGBS consortium
•
Novartis Institute for Tropical Diseases (NITD), Singapore
– Team leaders: Bryan Yeung, Zou Bin, Christophe Bodenreider
•
Genomic Institute of the Novartis Research Foundation (GNF)
– Team leaders: Arnab Chatterjee, Kelli Kuhen and Elizabeth Winzeler
•
Biomedical Primates Research Center (BPRC), Rijswijk (NL)
– Team leader: Clemens Kocken
•
Swiss Tropical Institute (STI), Basel (CH)
– Team leader: Matthias Rottmann
•
The NGBS (NITD
(NITD, GNF,
GNF BPRC,
BPRC STI) consortium
– Program Head: Thierry Diagana (NITD)
•
NGBS is funded by:
16 Chatterjee – Lecture at TSRI Oct 29 2014
GNF Malaria Cell-based Screening Summary
•
> 2M compounds screened at 1.25 M (3d7)
•
4851 hits < 1.2 M EC50 vs W2 and/or 3d7
– “Malaria Box” public resource for
malaria
l i research
h community
it
– https://www.ebi.ac.uk/chembldb/index.p
hp/compound
•
1,256 < 200 nM EC50 vs 3d7
•
>200 scaffolds represented in the
reconfirmed compound set
•
A much smaller subset active on liver-stages
g
of infection ~270 cmpds
How to best prioritize
chemistry
h i t resources?
?
Bl d t
Blood-stage
HTS reported
t d in:
i Plouffe
Pl ff and
d co-workers
k
PNAS 2008,
2008 9059
9059.
17 Chatterjee – Lecture at TSRI Oct 29 2014
Selection criteria for cell-based HLO at GNF
•
~5K reconfirmed blood-stage hits from HTS identified for further
evaluation were determined by
– Less than 1.25 µM EC50 with activity in 15 stain resistance panel
• Less than 5-fold potency shift between strains
– High selectivity in 6-cell line toxicity panel (SI > 20-fold)
– Novel chemotype for malaria
– Ease of synthesis (less than 7 steps)
– Good solubility,
solubility metabolic stability and limited CYP450 inhibition
– Multiple actives within a scaffold (if library diversity allows)
•
OPI algorithm used for clustering (developed at GNF) Yan, et al. J. Chem. Inf. Model.
2005 45,
45 1784 -1790
1790
•
Integrates SAR information into analysis of HTS data to prioritize interesting
singletons as well as lower activity compounds that have good SAR from the screen
Thi presentation
This
i will
ill focus
f
on 1 series
i prioritized
i i i d from
f
this
hi clustering
l
i
18 Chatterjee – Lecture at TSRI Oct 29 2014
Imidazolopiperazines Hit series
GNF-Pf-5069
EBI ID
HTS
W2
EC50
(µM)
GNF-Pf-5179
GNF-Pf-5466
HTS
3D7
EC50
(µM)
HTS
Huh7
EC50
(µM)
Powder
W2
EC50
(µM)
Powder
3D7
EC50
(µM)
Powder
Huh7
EC50
(µM)
GNF-Pf- 0.097
5069
0.063
>10
0.198,
0.473
0.161,
0.46
>100
GNF-Pf- 0.271
5179
0.235
>10
0.122
0.119
>79
GNF-Pf- 0.119
5466
0.116
>10
0.029
0.030
42.82
19 Chatterjee – Lecture at TSRI Oct 29 2014
Imidazolopiperazines Hit Series
• Strengths
– Unique scaffold for malaria, no biological activity
annotation with structure
– 4 step synthesis
– Veryy g
good solubility;
y; no activityy on Cyp450s
yp
– Not identified in >100 HTS screens performed at GNF
• Issues
– Moderate in vitro potency
– In vivo snapshot PK: poor oral exposure
• Likely due to catechol and glycinamide
– hERG inhibition (35% at 10 M, patch clamp)
20 Chatterjee – Lecture at TSRI Oct 29 2014
Compound Progression Scheme
MedChem Output
Falciparum Resistance Panel (15)
Protein binding shift
Human Serum shift of Pf EC50
PBFS <10
Huh7, HeLa, BaF3, CHO, 293T, HEp2
SI > 50
EC50 < 200nM; <5x shift
Tier I In Vitro ADME
Mouse snapshot PK
Tier II In Vitro ADME
Ex vivo
o in vitro
t o P berghei
be g e
 Solubility > 50 µM
 Low to moderate metabolism: hu, rat,
ms, dog, monkey microsomes
 CyP inhibition, 5 isoforms (>1 µM)
 Caco-2 > 3 x 10-6 cm/sec (A-B)
GNF
NITD/NIBR
Swiss Tropical Institute (STI)
Scaffold characterization of (selected analogs):
Liver-stage assays, Stage of action,
Rate of action
In vivo PK
MTD
Cmax >10x EC50, %F >20%, T½ >2-4hr, qd preferred
MTD (single-dose) > 5x efficacious dose
Safety and P/C profiling
PCP P5 Panel
Kinase/Protease Panels
In vitro tox (hERG, AMES, MNT)
yp induction and timeCyp
dependent inhibition
CPP assessment
Efficacy in vivo vs. P berghei
T
Transmission
i i and
d P vivax
i
assays
“Late lead” nomination
21 Chatterjee – Lecture at TSRI Oct 29 2014
Cell line Tox Panel (6)
3d7, W2, FCR3, FCB, HB3, Dd2, C188, NF54,
D6, D10, K1, CAMP-R, TM, 7G8, 3BA6




IC50 ≥ 5 µM (no prohibitive activity)
hERG potential (Patch clamp >10 µM)
Micronucleus: Negative
Ames: Negative
 Cure at fully tolerated dose, 3 day
dosing
 >99% reduction parasitemia
 ED50/90/99 measured
Hit-to-lead optimization Summary
H2N
O
N
N
N
NH
O
•
O
• Compound GNF776
GNF-Pf-5069
–
–
–
–
–
–
Pf 3D7 (EC50): 460 nM
hERG (binding) IC50 = 19 µM
Mouse metabolic stability (ER): 0.24
Solubility (pH 6.8): > 175 µM
PO Cmax (D.N.) = 16 nM / (mg/kg)
PO AUC0-5h
g g)
0 5h ((D.N.)) = 48.6 h*nM / ((mg/kg)
Poor exposure limited any further work
(*D.N.: dose normalized)
–
–
–
–
–
–
–
Pf 3D7 (EC50): 20 nM
hERG (binding) IC50 = 3.95 µM
CL = 60 mL/min/kg; Vss = 23.7 L/kg; T½ = 6.2 h
PO Cmax (D.N.) = 60 nM / (mg/kg)
PO AUC0-inf (D.N.) = 596.5 h*nM / (mg/kg)
F (%) = 87
Efficacy achieved in p berghei mouse model
Wu, et.al. J. Med. Chem. 2011, 5116.
Still need to improve hERG and potency
22 Chatterjee – Lecture at TSRI Oct 29 2014
Groebke-Blackburn Cyclization Route
P falciparum 3D7
EC50 = 30 nM
• 4 steps and 53% overall yield
• No column chromatography
23 Chatterjee – Lecture at TSRI Oct 29 2014
Imidazolopiperazine Core Optimization
8,8-dimethyl substitution significantly improves profile
• GNF179
• GNF268
–
–
–
–
–
–
–
–
Pf W2 (EC50): 9 nM
hERG (binding) IC50 = 6.6 µM
CL = 1179 mL/min/kg;
L/ i /k Vss = 106.76
106 76 L/k
L/kg
Oral T½ = 2.2 h
PO Cmax (D.N.) = 3.3 nM / (mg/kg)
PO AUCinf (D.N.) = 30 hr*nM
F (%) = 17
30x cross-resistance to GNF-Pf-5069 lab
evolved resistance strain
–
–
–
–
–
–
–
–
Pf W2 (EC50): 6 nM
hERG (binding) IC50 = 7.2 µM
CL = 21.92
21 92 mL/min/kg;
L/ i /k Vss = 11.8
11 8 L/k
L/kg
Oral T½ = 8.4 h
PO Cmax (D.N.) = 60.5 nM / (mg/kg)
PO AUCinf (D.N.) = 1035 hr*nM (mg/kg)
F (%) = 58
7x cross-resistance to GNF-Pf-5069 lab
evolved resistance strain
(*D.N.: dose normalized)
gem-Dimethyl group on piperazine improves potency and PK in mice
J. Med. Chem. 2012, in press.
24 Chatterjee – Lecture at TSRI Oct 29 2014
SAR Summary
• Optimization of the Imidazolopiperazine scaffold:
– Introduction of 8,8-dimethylgroup on piperazine improves in vitro and in vivo
potency by 10-fold
– Required glycine functionality is primary cause for hERG activity
Amino acid SAR
• Aryl required
•NH2 preferred; but also can use NHR, NR2,
OH and alkyl groups
• Para and meta substituted aromatic
preferred
• 1-2 carbon spacer to NH2 preferred
•Dimethylglycine improves in vivo PK
relative to glycine (mouse and rat)
• Ortho substituted aromatic not
preferred
H2 N
• Glycine least active on hERG (2-3 fold),
while all other amino acids in similar range
O
N
N
F
N
• Strong electron-withdrawing group
not preferred
NH
Piperazine ring SAR
• 8-substitution helps with potency and mouse
exposure (dimethyl > valine, alanine >
hydrogen)
• 5-dimethyl and 6-dimethyl moderately helps
potency, no real effect on PK
• Cyclic lactams (6-oxo)
(6 oxo) derivatives active with
8,8-dimethyl substitution and reduces
metabolism
25 Chatterjee – Lecture at TSRI Oct 29 2014
F
• Aniline required with unsubstituted
piperazine series
• Nitrogen required
required, but N-H
N H is not in
unsubstituted piperazines
• Variety of aromatics help potency (para and
meta preferred)
• Heteroaromatics and cycloaliphatics not
active in unsubstituted piperazine series, but
active with 8-dimethyl groups
• Can modestly effect hERG activity
hERG Optimization
Targeting the amine functionality
Modulate amine basicity:
GNF824
P. falciparum 3D7 EC50 = 50 nM
pKa = 3.3
Glycine amine pKa = 7.5
hERG (Dofetilide binding)
IC50 = 12.47 M
Solubility (pH 6.8)
6 8) = 34 mg/L
Replacement of amine with alcohol:
GNF711
P falciparum 3D7 EC50 = 300 nM
hERG (Dofetilide binding)
IC50 > 30 M
Solubility (pH 6.8) = 110 mg/L
GNF557
Optimize
P. falciparum
potency
P. falciparum 3D7 EC50
= 20 nM
Unable to progress due to poor metabolic stability and Log D
26 Chatterjee – Lecture at TSRI Oct 29 2014
SAR of the 3-position (aniline group)
pH 6.8
Solubilit
y (µM)
Rat
ER
hERG
(binding)
(µM)
Cmpd
X
R
Pf 3D7 EC50
(nM)
GNF156
NH
4 F Ph
4-F-Ph
10
>1000
0 52
0.52
64
6.4
GNF877
CH2
4-F-Ph
8
>1000
0.49
2.7
GNF767
O
4-F-Ph
3
>175
0.13
4.0
GNF765
S
4-Me-Ph
8
42
ND
4.2
GNF772
SO
4-Me-Ph
42
>175
0.37
26.0
GNF771
SO2
4 M Ph
4-Me-Ph
5210 (NF54)
ND
ND
ND
GNF713
SO
3-F-Ph
481
ND
ND
12.0
GNF714
SO
4-F-Ph
70
ND
ND
>30
GNF715
SO
3,5-Me2Ph
381
ND
ND
15.0
GNF716
SO
3,4-F2-Ph
89
ND
ND
27.0
Sulfoxide linkage is a good balance of potency, solubility and hERG
27 Chatterjee – Lecture at TSRI Oct 29 2014
Profile of Preclinical candidate: GNF156
•
•
•
•
In vitro Activity Profile of Imidazolopiperazines
– P. falciparum W2 potency: EC50 = 0.006 μM
– Compound exhibits < 5-fold shift across 15 strains
– Selectivity vs. Huh7 cells: >10,000 fold
– Complete transmission blocking at 500 nM
GNF156
In vivo Profile
– Good
G d orall PK/bi
PK/bioavailability
il bilit and
d exposure; T1/2 > 4 h
– In vivo efficacy in P. berghei mouse model: ED99 = 2.2 mg/kg
– Causal prophylaxis data: Fully protective at 20 mg/kg
ADME/Safety/Physicochemical Profile
– ADMET: solubility: >1000 µM @ pH1 and @ pH 6.8; no GSH trapping; no Cyp
induction or phototoxicity flags
– Clean PCP receptor panel, Ames, MNT and good metabolic stability
– Formulation suitable for tox studies identified (35x exposure multiples achieved in
rats and 16x in dogs)
– hERG activity at 6.6 µM in binding assay; 13 µM in automated patch clamp assay
Unique activity for blood/liver-stage
blood/liver stage infections and transmission blocking
28 Chatterjee – Lecture at TSRI Oct 29 2014
GNF156: Current Synthesis Route



29 Chatterjee – Lecture at TSRI Oct 29 2014
10 step route, but minimal purifications
27% overall yield
Low cost of goods for starting materials
GNF156: In vivo efficacy in P. berghei model
Compound
# of
animals
per dose
ED50 (mg/kg)
ED90 (mg/kg)
ED99 (mg/kg)
Chloroquine
5
1.9
5.2
8.4
Mefloquine
5
3.8
5.0
8.6
At
Artesunate
t
5
50
5.0
20 5
20.5
119
Imidazolopiperazine
GNF156
5
0.48
1.13
2.2
Oral Exposure in mouse at efficacious dose (2 mg/kg)
Cmax = 112 ± 5 nM
Tmax = 1 h
T1/2 = 4.8 ± 0.8 h
AUCinf = 945 ± 110 nM*h
%F = 32 ± 4
30 Chatterjee – Lecture at TSRI Oct 29 2014
GNF156 PK Properties in Mouse, Rat and Dog
GNF156 Single Dose PK in Mice
Mouse PK (5 / 20 mg/kg, n=3)
GNF156 Single Dose PK in Rat
Rat PK (3 / 10 mg/kg, n=3)
GNF156 Single Dose PK in Dog
Dog PK (0.5 / 2 mg/kg, n=3)
CL (ml / min•kg)
49 (mod)
CL (ml / min•kg)
24 (low)**
CL (ml / min•kg)
5 (low)
Vss (l/kg)
10 (high)
Vss (l/kg)
26 (high)
Vss (l/kg)
11 (high)
t1/2term.
(h)
1/2t
67
6.7
t1/2term.
(h))
1/2term (
4.9
t1/2term. (h)
28 5
28.5
AUC (h•nM) p.o.
12156
AUC (h•nM) p.o.
975
AUC (h•nM) p.o.
16142
Cmax (nM) p.o.
1539
Cmax (nM) p.o.
91
Cmax (nM) p.o.
553
Tmax p.o. (h)
1.0
Tmax p.o. (h)
1.6
Tmax p.o. (h)
1.7
Oral BA (%F)
74
O l BA (%F)
Oral
20
Oral BA (%F)
89
** Adjusted for RBC partitioning
• In rat whole blood, GNF156 showed moderate RBC partition with blood to plasma ratio of 3.5.
• GNF156 exhibits low to moderate blood CL,, high
g Vss, and reasonable bioavailability
y across species
p
• Exposure multiple of 16 achieved in dogs (AUCeff ~1000 h*nM at 2 mg/kg in mice). Long T1/2 in dogs.
31 Chatterjee – Lecture at TSRI Oct 29 2014
GNF156: Human efficacious dose projections
Species
Units
Mice
Rat
Dog
Human
Body weight
kg
0.025
0.25
8.7
70
Dose
mg/kg
5
3
0.5
1
Vss
L/kg
10.2
25.7
10.9
10.2 to 25.7
CL
mL/min/kg
49.2
84.4
4.8
0.9 to 4.5
T1/2
h
2
49
4.9
28 5
28.5
14 5 to 71
14.5
F
%
66
47
89
67
 Efficacy
y : ED99 in mice for KAF156 is 2.2 mg/kg.
g g
 Bioavailability : F was estimated to be 67% in humans
 Clearance : 0.9 to 4.5 ml/min/kg using various methods.
 Human efficacious dose (to show similar efficacy to the mouse 99% parasitemia reduction)
was predicted to be 2.3 to 48mg based on the following assumption
 Efficacy is similar in P.b infected mice and P.f infected human
 PK is similar in normal mice and P.b
P b infected mice
32 Chatterjee – Lecture at TSRI Oct 29 2014
Imidazolopiperazines Block Transmission
Novel activity for a potent blood-stage antimalarial
Stage V gametocyte transmission
mosquito feeding experimental procedure:
1. The compounds are added to a bloodmeal
with NF54 gametocytes
2. Six days after the fed mosquitoes are
dissected and checked for oocysts
3 Compounds can have an effect on
3.
mosquitoes, therefore the mortality of the
mosquitoes is checked every day
Control
GNF776
(0.7 M)
GNF776
(7 M)
% infected
mosquitoes
# Oocysts/ mosquito
95
49
60
1.5
0
0
Data furnished by Sauerwein lab
33 Chatterjee – Lecture at TSRI Oct 29 2014
Conclusions and path forward:
• GNF776 (lead cmpd) inhibits oocyst
development dose-dependently
• Optimized GNF156 blocks
transmission completely at 500 nM
• This novel activity is critical for
potential eradication goals
GNF156 preclinical safety summary
No significant tox at up to 30-45 (rat) or 135-250 (dog) fold over ED99 exposure

Genotoxicity


Mini Ames mutagenicity assay: negative
Phototoxicity

3T3 NRU assay: negative (PIF = 1.4)

2-week rat toxicity at 10, 30, and 100 mg/kg/day

100 mg/kg/day: salivation, respiratory distress,  lymphocyte counts and
WBC mild thyroid follicular hypertrophy/hyperplasia
WBC,
hypertrophy/hyperplasia, and possibly testicular
degeneration (1/5 rats but not in controls)

≥30 mg/kg/day:  platelets (17-20%), in 3/5 rats at 30 and all rats at 100
mg/kg/day, mild-mod  creatinine (up to 2-fold); mild (15%)  globulin

All doses:  CK (-16- 56%) and glucose (-6 - 15%) in several rats per
group

Dog rising dose with telemetry at 2, 10, 30, 100, 250 mg/kg

250 mg/kg: Transient heart rate increase without ECG changes
34 Chatterjee – Lecture at TSRI Oct 29 2014
Visceral Leishmaniasis
• Disease caused by kinetoplastids from Leishmania genus (L.
donovani, L. infantum / chagasi)
• Endemic to 82 countries with > 90% cases found in 3 regions:
- India / Bangladesh / Nepal
- Sudan / Ethiopia / Kenya
- Brazil
• Prevalence
- 500,000 new cases per year
- Mortality conservatively estimated as 100,000 deaths per year
• Routes of infection – through bite of Leishmania-infected sandfly
• Visceral Leishmaniasis natural history
- Most infections have sub-clinical character
- Incubation period typically 2-8 months
- clinical features : fever >80%, wasting > 70%, splenomegaly > 90%
(almost universal), hepatomegaly > 50%, anemia > 60%
y fatal,, cause of death usuallyy
- untreated disease almost always
concurrent illness (pneumonia, tuberculosis, dysentery)
• Current therapies
- Pentavalent antimonials, amphotericin B, paromomycin, miltefosine
p
amphotericin
p
B),
), toxicityy
- Drawbacks include cost ((liposomal
(antimonials, amphotericin B, miltefosine), drug resistance
(antimonials)
35 Chatterjee – Lecture at TSRI Oct 29 2014
Leishmania
Lih
i spp.
Extracellular
promastigotes
in insect gut
Sandfly
(Phl b t
(Phlebotomus,
Lutzomyia)
Regurgitated
R
it t d b
by iinsectt iinto
t
bite wound
Propagates
as amastigote
g
in vacuole of
macrophages
& dendritic
cells
Parasites transmitted back to
insect during blood meal
Selection of Hit Scaffolds from L. donovani HTS
Phenotypic, axenic amastigote HTS
(~700,000 compounds)
Lead Discovery Database (LDDB)
at GNF for integration HTS and med chem data
1,035 confirmed hits with EC50 less than 4 M
dose response heat map
BAF/Tel-AlK
BAF/Tel-Ros
BAF/Tel-InsR
BAF/Tel-IGF1R
BAF/Tel-TRKA
BAF/Tel-TRKB
BAF/Tel-TRKC
BAF/Tel-MET
BAF/Tel-MER
BAF/Tel-RET
BAF/Tel-RON
BAF/Tel-FGFR4
BAF/Tel-FGFR2
BAF/Tel-FGFR3
BAF/Tel-KDR
Column 22
BAF/Tel-FLT1
Column 19
BAF/Tel-FLT4
BAF/Tel-Kit
Column 16
BAF/Tel-FMS
Column 13
BAF/Tel-FLT3
Column 10
b
BAF/Tel-PDGFRb
Column 7
BAF/Tel-Tie1
Column 4
BAF/Tel-TIE2
Column 1
BAF/Tel-EphA3
wtBAF/
Confirmation axenic amastigote assay: 87/89 confirmed
BAF/Tel-EphB2
Clustering and purified compound order of representative scaffolds: 89 compounds
BAF/bRaf Pim1-1
BAF/Tel-Syk
BAF/Tel-ZAP70
BAF/Bcr-Abl
BAF/Tel-BMX
BAF/Tel-Lyn
BAF/Tel-Src
BAF/Tel-Lck
BAF/Tel-FGR
BAF/Tel-TYK2
BAF/Tel-JAK1
BAF/Tel-JAK2
High selectivity in Huh7 cell line toxicity (SI > 10-fold)
BAF/Tel-JAK3
3.7 1.5 2.6 3.7 1.6 0.9 3.3 0.8 1.8 2.3 7.0 3.0 4.4 4.2 4.4 3.7 2.8 3.2 2.4 4.4 5.4 6.1 3.9 2.6 1.9 2.8
3.5 3.7 3.7 2.5 2.1 1.9 1.4 4.3 4.8 2.6 3.5 3.4 0.5
No/ minimal annotation in > 60 HTS
Tier 1 ADME: microsome stability; Cyp inhibition; solubility: 43/89 tested
Preliminary Cost of Goods assessment
12 scaffolds for confirmation for activity in macrophage infection model
36 Chatterjee – Lecture at TSRI Oct 29 2014
Efficacy Comparison - Hits vs. Known Drugs
•
•
•
•
8 of 12 scaffolds show
activity
y < 10 
M on
intracellular amastigotes
2 of these 8 scaffolds have
no activity
y on axenic
amastigotes – fail to
reconfirm
4 of 12 scaffolds no activity
y
on intracellular amastigotes
at 50 M
Several interesting
g scaffolds
(on right) were not optimized
for chemistry despite
amastigote activity
Compound ID
Axenic Amastigote
IC50 (M)
Macrophage
infection
IC50 (M)
GNF6372
1.41
> 50
GNF6513
2.56
> 50
GNF7674
2.24
> 50
GNF3336
0.63
> 50
A h
Amphotericin
i i B
1 42
1.42
0 13
0.13
Miltefosine
G
GNF6372
GNF7674
37 Chatterjee – Lecture at TSRI Oct 29 2014
4.12
GNF6513
GNF3336
Leishmania Project
Current Hit-to-Lead Status
• 8 scaffolds have been identified with following properties
– Good SAR from screen (active and inactive members within a
scaffold)
ff ld)
– All commercially available compounds, not annotated with any antiinfective activity
– Preliminary in vitro ADME assessed, including:
• Liver microsome stability (mouse and human)
• CyP450 3A4 reversible inhibition (for DDIs)
• Solubility assessment in pH 6.8 and in simulated GI fluid buffer
(Fassif)
– ~1000
1000 compounds made on these series since Feb 2011
• Goal for project is to find 1-2 lead scaffold with in vivo
potency similar to miltefosine
38 Chatterjee – Lecture at TSRI Oct 29 2014
Triazolothiazoles hit GNF8023
Axenic HTS hit with moderate
macrophage activity
Unable
Unable to measure oral exposure
mouse plasma instability
Instability arises from cleavage of
the methyl ester
Series
S i h
has steep
t
SAR
SAR:
Acid compound (GNF5410) is
inactive on the parasite, along
with small amides ((total of 15
analogs made)
Other amines other than
piperidine also inactive
including azetidines,
azetidines
pyrrolidines and piperazines
Chemistry on hold for this series
39 Chatterjee – Lecture at TSRI Oct 29 2014
GNF8023
GNF5410
Potency
Macrophage Infection:
5.5 M
Solubility
HT-Eq Sol:
119 M
Microsomal Stability
ER Mouse:
0.97
Cyp Inhibition
CYP3A4:
4 M
Cytotoxicity
Mouse Macrophage:
>50 M
Mouse Model of Acute Visceral Leishmaniasis
L donovani propagation in BALB/c Mouse
L.
6
n = 5 animals per group
2 miltefosine doses used (10 mg/kg and 30 mg/kg)
p.o. dosing x 5 days (7 through 11 days)
liver parasitemia quantified on day 14 (qPCR)
4
Efficacy of miltefosine in BALB/c VL
1.0E+1
2
Liver parasitemia - fold change
e
Liver para
asitemia – fold cha
ange
tail-vein injection of amastigotes
parasitemia quantification in liver by histology and qPCR
Acute VL Model Validation
1.0E+0
0
0
7
14
Days Post-infection
Post infection
1.0E-1
1.0E-2
1.0E-3
1.0E-4
Vehicle
10 mg/kg
30 mg/kg
Treatment
Giemsa stain of liver imprints from infected BALB/c mouse
on day 14 post-infection. Infected macrophages containing
L. donovani amastigotes are indicated with arrows
40 Chatterjee – Lecture at TSRI Oct 29 2014
Results are in agreement with previously published
p
ED50 for miltefosine against
g
L. don.
data: reported
in mice: ∼ 6 mg / kg / day × 5 days)
Chagas disease
• Disease caused by kinetoplastid Trypanosoma cruzi
• Endemic to 18 countries in Central and South America
• Prevalence
- ~ 200,000 new cases per year
- Mortality
M t lit estimated
ti t d as 50,000
50 000 deaths
d th per year
• Routes of infection:
- 80% of infections attributed to bite of infected insect vector
- Blood transfusion
- Congenital transmission (4% of babies born to infected mothers)
- Orally via contaminated food
• Chagas disease natural history
- Acute phase: 7-9 days; often passed unnoticed; in 90% cases
acute disease resolves spontaneously
- Indeterminate phase: ~ 70% patients remain asymptomatic, risk of
sudden death due to cardiac abnormalities
- Chronic phase: develops 10-30 years after infection in ~ 30%
patients;
ti t leading
l di cause off congestive
ti heart
h t failure
f il
in
i Latin
L ti
America; costly treatment
• Current therapies
- Benznidazole, nifurtimox – nitroheterocycles with side effects
(dermatitis diarrhea,
(dermatitis,
diarrhea nausea,
nausea vomiting
vomiting, etc
etc.))
- Active on acute form of disease (80% cure rate), marginally on
chronic form (20% cure rate)
41 Chatterjee – Lecture at TSRI Oct 29 2014
T
Trypanosoma
cruzii
extracellular
epimastigotes
in insect gut
Kissing bug
(Triatominae
Ti t i
)
Transferred
T
f
d by
b insect
i
t feces
f
deposited near bite wound
propagates
intracellularly
in cytoplasm
as amastigote
invades many
different cell
types
Parasites are transmitted back
to insect during blood meal
Cross-Activity of HAT / VL Hits against T. cruzi
L. donovani
axenic amastigote
T. brucei
bloodstream form
410
377
93
214
322
175
Active
Cmpds
L. don. selective
25.8 %
T.brucei selective
23.7 %
L. don.+T.brucei
5.8 %
L.don.+T.cruzi
20.2%
T.brucei+T.cruzi
10.1 %
Pan-kinetoplastids
13.5 %
T t l T.
Total
T cruzii active
ti
44 7%
44.7%
L. don. HTS represents a more
abundant source of T. cruzi-active
compounds
d th
than T.
T brucei
b
i HTS
T. cruzii
T
8 day infection assay
42 Chatterjee – Lecture at TSRI Oct 29 2014
Unusually low overlap between L.
donovani and T. brucei hits outside
of p
pan-kinetoplastids
p
Identification of T. cruzi Amastigote Inhibitors
T. cruzi
8 day infection assay
(IC50 < 4 M)
T. cruzi
amastigote proliferation
(
(>75%
% ((50%)
%) inhibition @ 10 M))1
Subset of 574 L. don. / T.
brucei – active inhibitors tested
f activity
for
ti it att amastigote
ti t
proliferation stage1
148 compounds found active
(>75% iinhibition)
hibiti ) @ 10 M
M
concentration
409
132
(332)
(209)
16
(63)
~90% of amastigote stagespecific
ifi iinhibitors
hibit
were
previously identified as T. cruziactive in 8 day assay
Significant portion of T.
T cruzi
cr i –
active hits is likely to act at
another stage of parasite life
cycle
1data
obtained and provided by Juan Engel
Engel, Sandler Center
43 Chatterjee – Lecture at TSRI Oct 29 2014
T. cruzi Hit-to-lead Chemistry
N
• Hit-to-lead chemistry started Oct 2010
• All compounds from academic
collaboration compound collection at GNF
• 8 total scaffolds tested in footpad efficacy
model, developed by Rick Tarleton and coworkers
k
• Several scaffolds show good in vitro
potency and oral exposure, however no
efficacy in footpad model
N
H
N
O
GNF5063
GNF2609
Reasonable oral exposure
P
Poor
selectivity
l ti it on h
hostt cellll
Unable to remove chelating 2pyridine
Unable to seperate parasite
activity
ti it from
f
human
h
Cyp
C
Good oral PK
50% reduction in parasitemia
in footpad model
GNF8308
Poor selectivity on 3T3 host cell
Active on human Cyps (< 2 µM)
G d orall PK b
Good
butt no ffootpad
t d efficacy
ffi
44 Chatterjee – Lecture at TSRI Oct 29 2014
Cl
Chagas Mouse Footpad Model, Tarleton lab
• Efficacy model run in-house that employs a T. cruzi strain (CL Brener) expressing tandem
dimeric tomato red fluorescent protein (tdTomato)
p with 2.5 x 105 T. cruzi tdTomato trypomastigotes
yp
g
and
• Mice are infected in the hind footpad
the fluorescence signal intensity is measured as a surrogate of parasite load.
Vehicle
N
Day 6 post-infection
Day 8 postinfection
Day
D 6 post-infection
t i f ti
Day
D 8 post-infection
t i f ti
Day 11 postinfection
O
N
NHBn
O2N
Benznidazole
Day
D 11 postt
infection
Fluorescent tdTomato-expressing T. cruzi parasites imaged post-treatment
45 Chatterjee – Lecture at TSRI Oct 29 2014
Summary / Lessons Learned
•
Molecular weights greatly effect the cellular optimization
– Forces the chemistry to be very atom economical (high ligand efficiencies)
•
SAR can be challenging,
g g, but g
getting
g hits from HTS with g
good SAR is critical
– Better to have SAR than absolute potencies off the deck
– Potencies can be optimized relatively quickly compared to other parameters
•
•
•
•
•
Weekly cytotoxicity, protein binding shift potency, and cross-resistance data
to other scaffold lab-evolved resistance strains really allows for better
optimization of other parameters
MoA studies in parallel with optimization
St t with
Start
ith a simple
i l target
t
t profile
fil (e.g.
(
blood-stage
bl d t
antimalarial,
ti l i l th
then address
dd
liver-stages, transmission, etc)
Use extensive anti-target screening that has been developed over the last
decade due to focus on target-based
target based drug discovery
Academic collaborations have been very useful in using new assays for
medicinal chemistry optimization
46 Chatterjee – Lecture at TSRI Oct 29 2014
Acknowledgements (Kinetoplastids)
47 Chatterjee – Lecture at TSRI Oct 29 2014
Acknowledgements (Malaria Research)
Chemistry
• Advait Nagle
• Tao Wu
• Tomoyo Sakata
• Robert
R b tM
Moreau
• Jason Roland
• Pranab Mishra
• David Tully
• Valentina Molteni
Biology
• Kelli Kuhen
• Carolyn Francek
• Zhong
Zh
Ch
Chen
• Kerstin Henson
• Rachel Borboa
• James Gilligan
• Tae
Tae-gyu
gyu Nam
• Neekesh Dharia
• David Plouffe
• Case McNamara
• Stefan Meister
• Elizabeth Winzeler
48 Chatterjee – Lecture at TSRI Oct 29 2014
Pharmacology/Analytical
• Tove Tuntland
• Perry Gordon
• Jonathan Chang
• Matthew Zimmerman
• Liang Wang
• Todd Groessl
• Barbara Saechao
• Bo Liu
• Chun Li
• David Jones
• Wendy Richmond
• Kevin Johnson
• Tom Hollenbeck
• Lucas Westling
• Michael Kwok
• Tiffany Chuan
• John Isbell
Swiss Tropical and Public Health
Institute
• Matthias Rottmann
• Christoph Fischli
• Sonja Maerki
Informatics
• John Che
• Yingyao Zhou
GNF Management
• Jennifer Taylor
• Richard
Ri h d Glynne
Gl
• Martin Seidel
• Peter Schultz
NITD
• Bryan Yeung
• Zou Bin
• Anne Goh
• Suresh B. Lakshminarayana
• Veronique Dartois
• Thomas Keller
• Thierry
Thi
Di
Diagana
External collaborators
• Montip Gettayacamin (AFRIMS - Bangkok)
• Robert Sauerwein (NCMLS - The
Netherlands)
• Ian Bathhurst (MMV)
Acknowledgements (Malaria - Development)
Core team
Margaret Weaver
Xingmei Han
Giancarlo
Gi
l Francese
F
Sreehari Babu
Rita Ramos
Karen Beltz
Bo Han
Wen Shieh
Markus Baenziger
Novartis Tropical Medicines
Heiner Grueninger
Anne-Claire Marrast
Paul Aliu
Gilbert Lefevre
49 Chatterjee – Lecture at TSRI Oct 29 2014
Clinical team
Jens Praestgaard,
Praestgaard Ruobing Li
External support and advice
Marcel Tanner, Swiss Tropical Institute
Nick White, Oxford University