Discovery of inhibitors of MAPKAPK5,
a novel target for rheumatoid arthritis
Martin Andrews, PhD
Group Leader, Medicinal Chemistry
Inflammation 2010
18 November 2010
© Copyright 2010 Galapagos NV
Rheumatoid arthritis
 Common and difficult to treat disease
 chronic disease
 inflammation, pain & joint destruction
 leads to disability
 Affects ~ 1% of population
 5% of women over age 55
 increasing due to aging of population
 Large market size
 biologics: TNFα blockers (Enbrel® ,Remicade®, Humira®), IL-6 inhibitors
(Actemra® ), CD28 inhibitors (Orencia®), etc.
2009 global sales > €12 B
2
Target product profile
Novel drug based on Galapagos RA target
• Orally active drug for the disease
 once-daily dosing
• New way to treat RA
 reducing inflammation of the joint
 reducing destruction of bone & cartilage
• Better safety profile than existing therapies
3
Primary assay: target discovery
MMP1 assay principle
Disease relevant triggers
TNF
IL1
X
TNF
Y
Macrophage
Disease relevant cells:
RASFs
siRNA
Dominant negative
Compound
Disease hallmark
MMP1 expression
4
MAPKAPK5
Validation as RA target
Knockdown virus
Tissue
Target cell
Effect
Synovium
Synovial
fibroblast
Inhibition of TNF- /
IL1-induced MMP1 /
IL6 expression
Dominant
negative
5
Initial compound screening
Initial primary screening (10 µM)
BioFocus SoftFocus® collection
16,343 compounds
Retest compounds
468
IC50 determination
100
Hits triaged
58
3 series
1 series selected
6
Compound
Characterization of the hit series
O
N
HN
N
N
HO
N
H
N
O
OH
NH2
N
NH2
N
N N
H
Cl
IC50: 960 nM
O
OH
N
N
N
2,900 nM
NH2
N
320 nM
NH2
460 nM
Imidazopyrazines
Imidazopyridazines
Aminopyrazines
Aminopyridines
Potency




MMP1 inhibition


X

Selectivity

X
X
X
/ X
/ X
/ X
/ X
ADME
Imidazopyrazines selected as the lead series
7
R2 substitution in Series 1
Identification of suitable C5 functionality
O
O
N
HN
HN
N
N
N
HN
N
N
N
N
N
N
N
N N
H
N
H
O
pyrazole
IC50: 960 nM
O
N
O
NH2
4-carboxamidophenyl
1,300 nM
pyridone
390 nM
Functional groups with dual hydrogen bond donor/acceptor
capability are preferred at C5
8
Proof of Concept in RA
Prophylactic CIA study
100
Paw swelling (increase, mm)
60
40
TNF-
IL-6
Significant reduction
in disease-causing
cytokines
GM-CSF
vehicle
Compound A
vehicle
Compound A
vehicle
vehicle
0
Compound A
20
Compound A
% cytokine
80
Vehicle
Compound A
boost
p38 inhib
Methotrexate
IL-4
Minimal impact
on other cytokines
40 mg/kg, daily, p.o.
Compound A showed a positive effect on cytokine release
and disease development
9
Strategy in hit-to-lead
Series 1
N
N
HN
N
N
Limited C5-substituents
identified in hit-to-lead
had acceptable
MAPKAPK5 activity
H2N
Refine solubilizing group to
enhance ADME properties
• solubility
• microsomal stability
• plasma protein binding
N
Identify novel scaffolds – improve
further potency and IP position
O
Replacements to provide good PK
properties
• improve aqueous solubility
Addressing PK issues crucial for identifying pre-clinical candidate
10
Synthesis of Series 1
Br
Br
CO2Me
N
N
LiOH, THF
N
N
N
1
HN
N
N
reflux
Br
Br
R
N
2
iPr NEt, iPrOH,
2
N
Br
HN
R1NH
NH2
N
Br
Br
OEt
Br
1. Diphenylphosphoryl azide, Et3N
tBuOH, reflux
2. TFA/DCM
(62%)
N
MeOH, H2O
(quant.)
Br
CO2H
N
ArB(OH)2
R
N
N
DMF, H2O, NatOBu
Pd(PPh3)4, 90oC
1
N
Ar
11
MAPKAPK5 IC50 (nM)
Series 1 potency review
General potency of series improved over time
12
48% HBr Br
iPrOH
(59%)
OEt
Series 1 ADME/ PK
N
N
O
N
N
N
N
HN
HN
HN
HN
N
O
N
N
N
N
N
N
N
N
N
N
N
S
F
O
H2N
Aqueous solubility
(g/ml)
F
O
O
NH2
NH
N
H
O
<10
46
2
3.5
Not done
3.57, 11.2
1.73, 6.66
1.51, 12.2
9
20.8
90
Mouse hepatocyte
t1/2 >200min
AUC (ng.h/ml)
2,940*
321
2830
10900
Cmax (ng/ml)
591*
169
1350
2270
Caco2
(ER, Papp A>B
(cm.10-6sec-1))
Microsomal stability
(t1/2, mouse)
Mouse PK: 5 mg/kg, p.o.
* 20 mg/kg, p.o.
13
MAPKAPK5 homology model
Postulated binding mode for Compound D
• MAPKAPK5 homology model based on
MAPKAPK2 X-ray crystal structure 1NXX.PDB
• Scaffold nitrogen N2 and anilino NH at
C8 interact with hinge (Met174)
• Amido NH on furan can form hydrogen
bond with Glu131 on C-helix
 alternative conformations available
which may enable interaction with side
chains of Asp218 or Lys120
• t-Butylpiperazine occupies solubilizing
sub-pocket
Asp218
Lys120
Glu131
Met174
Glu172
Met171
14
Identification of a novel scaffold
• Alternative scaffold types proposed and prioritized
 in silico analysis, IP status, synthetic tractability
• Representative compounds from 8 chemical series synthesized
O
O
N
N
HN
HN
N
N
N
N
N
N N
N N
H
N N
H
Imidazopyrazine
Series 1
IC50: 960 nM
Triazolopyrazine
Series 2
810 nM
Triazolopyrazines selected as second lead series
15
Synthesis of Series 2
Br
Br
CO2Me
N
LiOH, THF
N
N
OMe
Br
Br
Br
N
N
NH2OH.HCl
N
O
N
B
O
PPA, 70oC
2h
(67%)
N
O
N
HN
Br
HN
N
N
N N
N N
H
N N
H
DMF, H2O, NatOBu
Pd(PPh3)4, 90oC
(36%)
N
N
H2N
N N
iPr NEt, iPrOH,
2
Br
(94%)
16
OH
Br
O
O
N
N
(90% over 2-steps)
Br
H
N
N
MeOH
N
EtOH
NH2
N
N
Br
OMe
Br
1. Diphenylphosphoryl azide, Et3N
reflux
tBuOH,
2. TFA/DCM
(62%)
N
MeOH, H2O
(quant.)
Br
CO2H
N
N
N
reflux
N N
Br
MAPKAPK5 IC50 (nM)
Series 2 potency review
Created Date
General potency increased over series 1
17
SAR Summary
Series 2
Para- substituted aryl most effective, piperazinyl substitutions
showed best activity, ADME properties
Pyridines in place of phenyl group also potent, but poorer ADME/PK
Other alternatives less effective
Other linkers, meta- substitutions less potent or poorer ADME/PK
bis- substitutions gave increased selectivity, but worse ADME
H-donor needed
for potency
R1
H N
N
Loss of activity with
R2 in 6-position
Deletion analogue inactive;
2- bromo compound 10
fold less active, 2-Me
analogue inactive
N
N N
R2
Good activity seen only with H-bond donor + acceptor groups present
Amides most effective, little impact on PK
18
SAR in the furan-carboxamide series
Phenyl ring is not crucial for
MAPKAPK5 activity
O
Basic substituents are
well tolerated
N
N
N
HN
S
IC50: 70 nM
80 nM
160 nM
2-Carboxamido-4-furyl gives
optimal in vitro and PK profile
N
N
O
N
IC50: 40 nM
O
H2N
N
N
N N
70 nM
O
O
N
N
O
IC50 180 nM
IC50: 14 nM
H2N
IC50: 40 nM
O
IC50: >10,000 nM
19
R2 modification PK
Series 2
O
O
HN
HN
N N
N N
N N
H
IC50 (nM) MAPKAPK5
Plasma protein binding (%)
• human; mouse
AUC (ng.h/mL)
Cl (mL/min/kg)*
T1/2 (h)*
F (%)
N
HN
N
N
N
N
N
N
N
N
N
N N
O
H2N
O
O
N
H
810
180
100
89.5; 85.5
89.5; 87
93; 94.5
2,090
1,234
1,777
27
55
23
0.65
4.1
1.4
68
82
50
Modified C5-substituents provide good PK properties
Mouse PK: 1 mg/kg i.v. (PEG400/H2O)
* 5 mg/kg p.o. (0.5% methyl cellulose)
20
Series 1 and 2
Phenotypic assays
Compound
MAPKAPK5
IC50 (nM)
MMP1
EC50 (nM)
MMP13
EC50 (nM)
B
38.7
7740
784
C
32.9
1980
417
D
37
990
320
E
36.8
769
315
F
19.8
1450
581
G
29.1
1150
259
21
Assay cascade
Kinase assay Phenotypic assay
Cytotoxicity
Selectivity
1,168 compounds
ADME assays
375
PK (rat, mice)
97
CIA model
35
6
Dose response CIA
Dog PK
4
PCC selection
22
Compound D profile
Parameter
Compound D
MAPKAPK5 potency (IC50, nM)
N
N
HN
N
N
N N
O
H2N
37
Cell assays (EC50, nM)
• MMP1
• MMP13
Human microsomal stability (t½, min)
• human; rat; mouse
Hepatocyte stability (t½, min)
• human; rat; mouse
Plasma protein binding (%)
• human; rat; mouse
CYP inhibition (µM)
• 3A4, 2D6, 2C9, 2C19, 1A2
O
Compound D
hERG binding (µM)
87; 84; 87
All >10
>10
PK (5 mg/kg p.o.)
• Mouse AUC (ng.h/mL)
• Rat AUC (ng.h/mL)
3,769
2,763
TNF- in serum
150
100
Vehicle
*
SB-203580
* *
Compound D
TNF- a (% of vehicle)
>130; >200; >200
1,400
Mouse LPS model
0
All >90
Therm. solubility (pH 3; µM)
23
50
990
320
30 mg/kg, p.o.
Mean +/- SEM, unpaired students t-test, n=6
* p<0.05 versus Vehicle
24
mAb model
Clinical score and swelling
75
0
2
4
6
***
***
0
8
Ve
hi
cl
Days after start dosing
MAB
LPS boost
AUC paw swelling
1.0
0.5
Compound D
2.5
**
0.0
0
2
4
6
Days after start dosing
Ve
hi
cl
-0.5
dex
e
0.0
8
Mean +/- SEM, Students unpaired t-test
* p<0.05, **p<0.01, ***p<0.001 versus vehicle
Vehicle
5.0
***
Dosing 40 mg/kg, p.o.
25 n=12 vehicle, n=8 treatment groups, n=5 dex
PK dose proportional in mice
 PK dose proportional within 3-30 mg/kg dose range
Cpd D (mg/kg)
Cmax
(ng/mL)
Tmax
(h)
3 mg/kg
10000
3
10
30
188
1453
5243
1
3
3
AUC(0-24h)
(ng.h/mL)
1242
9726
38170
AUC(0-)
(ng.h/mL)
1252
9758
nr
T1/2
(h)
3.6
2.8
nr
Cmax/dose
63
145
175
AUC(0-24h)/dose
414
973
1272
AUC(0-)/dose
417
976
nr
10 mg/kg
30 mg/kg
1000
Cpd D (ng/mL)
change in paw
swelling (mm)
1.5
de
x
0
25
D
2
D
4
50
Cp
d
6
Cp
d
8
de
x
AUC clinical score
change in
clincal score
LPS boost
10
e
MAB
12
100
10
1
0
nr: not reported
4
8
12
Hours
26
16
20
24
Dose-response CIA model
AUC  clinical score
 Clinical score
Clinical score
4.5
3.5
2.5
1.5
0.5
-0.5
-1.5
-2.5
0
2
4
6
8
10
12
14
25
15
**
5
*
**
***
-5
-15
Vehicle
-25
Enbrel 10mg/kg/3d (ip)
Cpd D 3mg/kg/d (po)
Days after start treatment
AUC paw thickness
0.5
0.0
Cpd D 10mg/kg/d (po)
7
Cpd D 30mg/kg/d (po)
5
3
1
**
**
***
***
-1
-3
-0.5
-5
0
2
4
6
8
10
12
14
-7
Days after start treatment
27
Dose-response CIA model
micro CT
5.00
4.50
4.00
3.50
Vehicle
3.00
*
2.50
2.00
1.50
1.00
Cpd D 10 mg/kg/d po
0.50
Ve
hi
cl
e
0m
g/
kg
/3
Cp
d
d
ip
D
3m
g/
kg
Cp
/d
d
po
D
10
m
g/
kg
Cp
/d
d
po
D
30
m
g/
kg
/d
po
0.00
Cpd D 30 mg/kg/d po
En
br
el
1
-1.0
Number of objects per slice
Increase in paw
thickness (mm)
1.0
28
Dose-response CIA model
Histology total score
14
°°°
RA total score
12
cell
infiltration
bone &
cartilage
lesion
Anova unpaired t-test
***p<0.005 vs CIA-vehicle
°°°p<0.005 vs Intact
pannus
10
8
6
***
***
Vehicle
***
4
2
0
INTACT
CIAvehicle
Enbrel
Cpd D
Cpd D
Cpd D
10mg/kg 3mg/kg 10mg/kg 30mg/kg
(ip) 3x week (po)
(po)
(po)
Cpd D
Compound D displayed significant RA sparing at 10 mg/kg
29
GLPG0259 pre-clinical summary
• Novel mechanism of action
• Differentiation from competitor RA candidate drugs in
development
• Orally active at 1 mg/kg/d in CIA model of RA
• Bone protection effects in CIA after oral dosing
• Nominated as a pre-clinical candidate in 2008
• Successfully completed toxicity package
• Entered Phase I clinical trials in 2009
30
Fluorescent molecular tomography
Molecular imaging of RA
• MMP/cathepsin probe injected into mouse
• Probe becomes activated once cleaved by
joint degrading enzyme
excretion
intact
MMPSense680
Similar results with
ProSense 680
Vehicle p.o.
CIA diseased
GLPG0259
4-day treatment
1 mg/kg p.o.
GLPG0259 reduces level of joint degrading enzymes
in mouse RA model
31
Multiple ascending dose study
Methotrexate (ng/mL)
Interaction with methotrexate
50 mg
Cmax (ng/mL)
Tmax (h)
AUC0-24h (ng.h/mL)
GLPG0259
GLPG0259
+ MTX
44.4 (21)
45.8 (26)
4 [2-7]
5.5 [2-7]
849 (25)
870 (27)
MTX Day 7
MTX + GLPG0259 Day 14
MTX
MTX
+ GLPG0259
Cmax (ng/mL)
139 (19)
114 (21)
Tmax (h)
1.5 [1-2]
2 [1-4]
AUC0-24h (ng.h/mL)
623 (21)
652 (15)
No interaction between GLPG0259 and MTX
32
GLPG0259 Phase I clinical trial
Exposure and interaction studies
 In vivo (CIA model) activity exposure levels exceeded
 No interaction between MTX and GLPG0259
MTX plasma levels
(ng.h/mL)
GLPG0259 plasma levels
CIA model activity
25 mg
50 mg
50 mg
75 mg
+ 7.5 mg
MTX
MTX
MTX
+ 50 mg
GLPG0259
33
GLPG0259 Phase I
Summary
• Good safety
 no effect on vital signs, cardiovascular safety, lab chemistry & hematology
 most common adverse effects (GI-related) easy to monitor
 maximum tolerated dose: 50 mg q.d.
• Good PK profile
 supports once-daily oral dosing
 similar safety & PK profile
 feasible to combine with MTX in Phase II studies
 capsule formulation tested/selected for Phase II trial
Phase II study started October 2010
34
Towards a novel drug in RA
Discovered target MAPKAPK5 using RA patient cells
(protease MMP1 as disease marker)
Identified compounds that inhibit MAPKAPK5
GLPG0259 effective in mouse RA model
(reduces inflammation and bone/joint destruction)
Target discovered in
disease relevant human
cells
Increased chance that
GLPG0259 will be effective
in RA patients
GLPG0259 reduces joint degrading proteases
in mouse RA model
GLPG0259 effective in RA patients?
(Phase II trial started Oct 2010)
35
Acknowledgements
Biology
Chemistry
Analytical Chemistry
Reginald Brys
Veerle Coose
Graham Dixon
Richard Janssen
Angelique Le Coz
Giocondo Lorenzon
Kevin Nash
Line Oste
Philippe Pujuguet
Cynthia Saint Marc
Andrew Self
Nick Vandeghinste
Martin Andrews
Gregory Bar
Veronique Birault
Mark Chambers
Andrew Clase
Stephen Fletcher
John Harris
Kim Hirst
Dan Hookins
William Kofie
Geoff Lawton
Angus MacLeod
Olga Roussel
Mike Russell
Wolfgang Schmidt
Alex Sudau
Pete Thomas
Giovanni Tricarico
Rawl Hardial
Rebecca Noble
ADME/PK
Ellen van der Aar
Alan Beresford
Kelly Dong
Nick Foster
Florence Namour
Evi Narinx
36
Computational Chemistry
Vivienne Allen
Joelle Lee
Clinical
Johan Beetens
Stan Blockhuys
Marc De Weer
Frédéric Vanhoutte
Piet Wigerinck