CRTh2: Can residence time help?

CRTh2: Can Residence Time help ?
25th Symposium on Medicinal
Chemistry in Eastern England
Fielder Centre,
Centre
Hatfield, Hertfordshire, UK
Thursday 24th April 2014
Rick Roberts
Almirall
Barcelona, Spain
Introduction
Paul Ehrlich, The Lancet (1913), 182, 445
“A substance will not work unless it is bound”
100 years on:
“how
how long it’s
it s bound determines how it works”
works
2
Energetic concept of Residence Time
Standard model
kon
[R] + [L]
unbound
[RL]
koff
bound
Energy
Potency is determined by the difference
between two rates
(R·L)‡
On rate
kon
Off rate
koff
[R] + [L]
SAR
koff
per second
kon
per second · per concentration
Potency = concentration
Potency
Ki
[RL]
3
Ki =
Equilibrium binding assays
measure potency but ignore kinetics
Energetic concept of Residence Time
Standard model
kon
[R] + [L]
[RL]
koff
unbound
bound
Energy
Structure Kinetic Relationships (SKRs)
h
have
b
been llargely
l iignored:
d
Residence Time
Dissociation half-life
Slow or fast kinetics
Off-rate, koff
(R·L)‡
SKR
On rate
kon
(R·L)‡
Off rate
koff
The binding kinetics are controlled by the
transition state energy
slow associating compounds that are slow dissociating
fast associating compounds that are fast dissociating
[R] + [L]
Potency
Ki
Are all equipotent
[RL]
If we can control the transition state energy, we can control the binding kinetics
4
What is the transition state (R·L)‡ ?
Standard model
kon
[R] + [L]
unbound
[RL]
koff
Solvated
Ligand
[L]
Solvated
Receptor
[R]
“Home”
Home
[R] + [L]
bound
Energy
Fleeting existence
(R·L)‡
Protein conformational
changes?
“Journey”
Entropy?
Partial
desolvations
?
(R·L)‡
Partial
electrostatic
interactions?
[R] + [L]
Potency
Ki
[RL]
Water molecules
Molecular interaction points
5
“Destination”
[RL]
A physical
process
What is the transition state (R·L)‡ ?
Standard model
kon
[R] + [L]
unbound
[RL]
koff
Solvated
Ligand
[L]
“Destination”
Destination Solvated
Receptor
[R] + [L]
[R]
bound
Energy
Fleeting existence
(R·L)‡
Protein conformational
changes?
Entropy?
Partial
solvations?
“Journey”
(R·L)‡
Purely
qualitative
[R] + [L]
Potency
Ki
[RL]
Water molecules
Molecular interaction points
6
“Home”
[RL]
Partially
break
electrostatic
interactions?
A physical
process
Why bother to control Binding Kinetics ?
Residence time / Dissociation half-life
minutes
hours
days
level
seconds
PD
PK
time
Mechanistic
Surmountable
Insurmountable
Pharmacodynamic
Potency has little influence over these behaviours
Partial vs Full agonism
 M3 agonists
 A2A agonists
7
Efficacy vs substrate concentration
 Lovastatin
Candasartan
Mechanism-based toxicity
p
Haloperidol
p
 Clozapine
 Celecoxib
Aspirin
Duration of Action
 Ipratropium vs Aclidinium
Kinetic Selectivity
 M3 vs M2 antagonism
The CRTh2 Programme
8
Brief introduction to CRTh2
Real name:
 Chemoattractant Receptorhomologous molecule
expressed on T-Helper 2 cells
 Also known as DP2
CRTh2 activation:
 induces a reduction of intracellular cAMP
and calcium mobilization.
 is involved in chemotaxis of Th2
lymphocytes, eosinophils, mast cells and
basophils.
CRTh2 and DP1 review. Nat. Rev. Drug Disc. (2007), 6, 313
 inhibits the apoptosis of Th2 lymphocytes
CRTh2 antagonism:
 stimulates the production of IL4, IL5, and
IL13, leading to:
 eosinophil
i
hil recruitment
i
and
d survival
i l
 mucus secretion
 airway hyper-responsiveness
 immunoglobulin E (IgE) production
 etc
t
 Should block pro-inflammatory PGD2 effects on
key cell types
9
 Potential benefit in:
 asthma
 allergic rhinitis
 atopic dermatitis.
Why Long Residence Time in CRTh2 ?
Indole acetic acids
HO
O
N
N
?
F
Oxagen-Eleventa
OC-459
Phase III
AstraZeneca
AZD-1981
Phase II
Actelion
Setipiprant
Phase III
Novartis
R = CF3 QAV-039
R = H QAW-680
Phase II/II
Merck
MK-7246
Phase I
Pulmagen-Teijin
ADC-3680
Phase II
Aryl acetic acids
?
Indomethacin
Weak agonist
Phenoxyacetic acids
AstraZeneca
AZD 5985 or
AZD-5985
AZD-8075
Phase I/I
10
Boehringer
g
BI671800
Phase II
Panmira
AM461
Phase I
Array
y
ARRY-502
Phase II
Panmira
AM211
Phase I
Amgen
AMG-853
Phase II
Roche
RG-7581
Phase I
Our internal programme
We want to find a once a day, oral, low dose (≤ 10 mg) CRTh2 antagonist for mild-moderate asthma
All antagonists are acids.
Acids generally have poor PK
Acid
Zwitterion
Human PK profiles of
150 carboxylic acids
Adapted from Obach,
Drug.Metab.Disp.
(2008), 36, 1385
PK Half-life vs Volume
Volume of d
V
distribution
n (L/kg)


10
Ideal
CRTh2
zone
1
Total body
water volume
Total plasma
volume
0.1
0
6
12
18
24
Human terminal half-life (h)
Clinical dosing of first CRTh2 antagonists – high dose and twice daily

Therefore, we chose to deliberately look
for slowly dissociating CRTh2 antagonists:



11
level

PD
To maintain receptor occupancy beyond
normal PK and extend duration of action
to reduce the pressure of finding a
carboxylic acid with desirable PK properties
to add the possibility of extra protection
due to insurmountability against PGD2 burst
PK
time
Insurmountable
Pharmacodynamic
Are there slow-dissociating CRTh2 antagonists?
Ramatroban
TM-30642
TM-30643
TM-30089
(CAY-10471)
Structure not
disclosed
Merck
MK-7246
Compound
PGD2
Potency
Dissociation
half-life*
Amira
AM432
Pulmagen
ADC-3680
Reference
Ramatroban
pA2
36 nM
5 min
TM-30642
TM
30642
pA2
20 nM
8 min
--- / / ---
TM-30643
pA2
4 nM
(12.8 years)
--- / / ---
TM-30089
pA2
(13.5 years)
--- / / ---
MK 7246
MK-7246
Ki
2 5 nM
2.5
33 min
Mol Pharmacol.
Mol.
Pharmacol (2011),
(2011) 79,
79 69
PGD2
KD
11 nM
11 min
Bioorg. Med. Chem. Lett. (2011), 21, 1036
AM432
IC50
6 nM
89 min
ADC-3680
Ki
1.6 nM
20 min
12
Mol. Pharmacol. (2006), 69, 1441
--- / / --American Thoracic Society (ATS), May 17-22, 2013
In vitro dissociation assays CRTh2
Surmountability vs Insurmountability
PGD2 agonist
dose-response curve
Re-mountable inhibition
6000
Rea
adout
Readout
15 min
TM 30089
TM-30089
0 → 5 µM
Readout
24 h
4000
2000
TM 30089
TM-30089
(CAY-10471)
0 1 nM
0.1
10 nM
1 µM
PGD2 Concentration
15 min < TM-30089 Dissociation half-life < 24 h
Surmountability is just a question of time
in vitro washout Dissociation assay
TM-30089
at 10 × IC50
Then PGD2
at 10000 × EC50
13
“Worst case” scenario for
dissociation half-life

N rebinding
No
bi di possible.
ibl

Physiological system may
be less demanding
Mean In
nhibition %
100
80
Dissociation t1/2 ~ 80 min
60
40
20
0
0
4
8
12
16
Time (h)
20
24
28
100 µM
Pyrazoles
 Indole nucleus developed by Oxagen. Beginnings of slow dissociation observed
 First series of Pyrazoles gave active compounds, but no significant residence time
(BMCL, 2013, 23, 3349)
 Second series of Reverse Pyrazoles gave a similar story
(Eur J Med Chem, 2014, 71, 168)
Indole
Pyrazoles
Reverse Pyrazoles
Ph
General
Oxagen
X = CH
X=N
General
GTPS IC50 (nM)
14
7
4
4 – 100
35
32 – 450
Dissociation t½ (h)
1.3
0.2
1.9
0.02 – 0.7
0.2
0.04 – 0.7
SAR pretty good. No SKR advances observed in either core or tail sections
Pyrazoles series abandoned for general lack of Residence time
14
Pipas
 Third series of Pyrazolopyrimidinones (Pipas) gave active compounds with long residence
(manuscript in preparation)
 Core SAR flat. Core SKR varied
Pyrazolopyrimidinones (Pipas)
*
*
*
*
*
*
*
*
*
*
*
*
Indole
Oxagen
PiPa
N-Me
N-Bn
diMe
N-Me*
N-CHF2*
GTPS IC50 (nM)
14
5
5
1
4
5
2
Dissociation t½ (h)
1.3
2.3
5.3
6.6
10
8
21
15
Pipas
 Tail SAR flat. Tail SKR varied
 Sulphone positioning ultimately affects SKR
O
HO
O
S R
O
N
Ortho substitution
N
H
O
Para substitution
R1
R2
GTPS IC50
Dissociation t½
R
GTPS IC50
Dissociation t½
H
H
5 nM
2.3 h
Ph
170 nM
n.d.
MeO
H
4 nM
4.2 h
Me
900 nM
n.d.
H
F
5 nM
33h
3.3
B
Bn
3 nM
M
09h
0.9
MeO
F
7 nM
23 h
16
Ultimately Potent but no duration

Pipa series was essentially impermeable.

Not suitable for an oral program

Still, a valuable tool to validate the PK-PD disconnection in guinea pig
PK-PD Disconnection Model
Indole Eosiniophilia PK-PD
PK PD
Pre-administration
time
1h
F-Indole guinea pig data
Dissociation t½ 1.3 h < PK t½ 5.3 h
 Inhibition of eosinophilia (PD) is
purely dependant upon systemic
levels (PK)
 No PK-PD disconnection
30
100
3
3
24 h
0.3
50
0.3
0
0.03
10
Dose mg/kg
100
1000
10000 100000
Plasma conc ((ng/ml)
g )
Pre-administration
time
3
3
0.3
diMe-Pipa guinea pig data
Dissociation t½ 20 h >> PK t½ 0.9
09h
 Inhibition of eosinophilia (PD) outlasts
drop in systemic levels (PK) at 17h
after dosing
 PK-PD disconnection (hysteresis)
17
0.3
0.03
Dose mg/kg
0.003
PK-PD Disconnection Simulations
PK half-life 1 h
“PD”
GPCR threshold
A 1 h Dissociation half-life is barely
y noticeable in the PD
Poor drug – short duration of action
A 12 h Dissociation half-life is definitely observable
A short acting drug keeps working long beyond its
expectations
t ti
At t = 0,
[Ligand] = 10×IC50
PK half-life 12 h
“PD”
GPCR threshold
A 1 h Dissociation half-life goes completely unnoticed
Good PK means once-a-day dosing
A 12 h Dissociation half
half-life
life is barely noticeable
A great drug is really efficacious over 24 h
At t = 0,
[Ligand] = 10
10×IC
IC50
Next day
Next dose
For the greatest observable effect, Dissociation half-life >> PK half-life
18
For a recent article, see Drug Disc. Today (2013), 18, 697
How long is long residence ?
Saturate Receptor,
then washout
1st half-life
2nd half-life
3rd half-life
Total receptor
population
Ligand lost
Too dilute to rebind
Cleared from circulation
Receptor Occupancy by exponential decay
Re
eceptor Oc
ccupancy
100
80
An antagonist fully saturating a GPCR
will lose robust efficacy after about
half a Dissociation Half-life
Receptor Occupancy required
for a GPCR antagonist.
60
40
A twice-daily
twice daily compound achieves 12 h PD
coverage due to PK levels
20
0
0
1
2
3
4
6
8
Half lives of dissociation
Half-lives
19
Pharmacol. Therap. (2009), 122, 281-301
10
To turn a twice-daily compound
into a once-daily compound,
we want to add on a Dissociation Half-life of ≥ 24h
Molecular Determinants of Long Residence:
Structure Kinetic Relationships (SKR)
20
Where is long Residence?
Selected lit reports or Structure Kinetic Relationships (SKRs)
 Trend analysis of D2 antagonists. Bioorg. Med. Chem (2011), 19, 2231.
 Trend analysis of Pfizer and literature data. Med. Chem. Comm. (2012), 3, 449
 Review of molecular determinants
determinants. Drug Disc.
Disc Today.
Today (2013),
(2013) 18,
18 667
Are Bigger compounds longer resident?
Energy
Molec
cular Weig
ght
(R·L)‡
[R] + [L]
[RL]
Reaction coordinate
Nope. Not diffusion controlled
No “Kinetic Efficiency”
21
Molecular size / weight ?
Lipophilicity ?
Charged state ?
Don’tt know ?
Don
Where is long Residence?
Are more potent compounds longer resident ?
Energy
(R·L)‡
[R] + [L]
More
potent
Longer
resident?
“Non” resident
10 – 1000 nM
[RL]
[RL]
Reaction coordinate
diffusion limit
Intuitive, but not that helpful
We want the most potent compounds anyway
22
Long residence
1 – 100 nM
Where is long Residence?
Are more polar molecules longer resident?
(R·L)‡
Energy
Is desolvation
relevant here?
Dissociation Half-life vs TPSA
(R·L)
(R
L)‡
TPSA (Å2 )
T
150
[R] + [L]
[RL]
Biaryl
Pyrazole
Pipa
6 Mem
Competitor
Other
100
50
0
Short Residence
1
2
4
8 16 32
Reaction coordinate
Dissociation Half-life (h)
No.
Freedom to adapt polar surface area to modulate physicochemical properties
23
Where is long Residence?
Are Impermeable compounds are Long Resident in CRTh2 ?
PAMPA
A
Permeability (x
x10-6 cm/s)
ResTime vs Permeability
100
10
Biaryl
Mono
Other
Pyr
RevPyr
PiPa
Std
1
0.1
0.01
0.001
0.0001
0.03125
0.0625
0.1250.25 0.5
1
2
4
8
Di
Dissociation
i ti H
Half-life
lf lif (h)
Not Intuitive
24
16 32
Where is long Residence?
Are Impermeable compounds are Long Resident in CRTh2 ?
PAMPA
A
Permeability (x
x10-6 cm/s)
ResTime vs Permeability
100
10
Biaryl
Mono
Other
Pyr
RevPyr
PiPa
Std
1
0.1
0.01
0.001
0.0001
0.03125
0.0625
0.1250.25 0.5
1
2
4
8
Di
Dissociation
i ti H
Half-life
lf lif (h)
Not Intuitive
And ultimately not true
Initially looking at an incomplete picture
Design permeable compounds for oral delivery
25
16 32
Where is long Residence?
Are more lipophilic compounds longer resident?
Energy
Dissociation Half
Half-life
life vs cLogP
(R·L)
(R
L)‡
Biaryl
Pyrazole
Pipa
6 Mem
Competitor
Other
6
cL
LogP
[R] + [L]
[R] + [L]
More
potent
[RL]
Reaction coordinate
Probably just pushes up
energy of [L] in free water
4
2
0
Short Residence
1
2 4 8 16 32
Dissociation Half-life (h)
Lipophilic molecules are often more potent,
because once bound,, the molecule doesn’t want to go
g back into the surrounding
g water
26
Where is long Residence?
Are more lipophilic compounds longer resident?
Within particular sub-series, there was often a relationship
Dissociation Half
Half-life
life vs cLogP
7-Azaindole
21 nM
IC50
Diss t½ 5 h
cLogP
-0.6
6
cL
LogP
CF3-Indole
I d l
37 nM
IC50
Diss t½ 27 h
cLogP
5.9
isoxazoles
4
2
0
Short Residence
1
2
4
8 16 32
Dissociation Half-life (h)
5-fold increase in Dissociation half-life
For a 1,000,000-fold increase in lipophilicity
Workable in final optimisation, but not without its associated risks for oral delivery
27
Where is long Residence?
Series by series
First example of series
6-Mem Series
10
5
0
5
4
3
2
1
0
1 min
1
2 4
8 16 32
1
2 4
8 16 32
Dissociation Half-life (h)
Dissociation Half-life (h)
Pipa Series
Biaryl Series
10
8
6
4
2
0
5
4
3
2
1
0
1 min
Number of Compounds
Number of Compounds
Competitor Compounds
Nu
umber of Compounds
15
Nu
umber of Compounds
Nu
umber of Compounds
Pyrazole Series
1 min
1
2 4
8 16 32
Dissociation Half-life (h)
60
40
20
0
1 min
1
2 4
8 16 32
Dissociation Half-life (h)
1 min
1
2 4
8 16 32
Dissociation Half-life (h)
 Chemical series drives Residence Time: If you’ve got it, you’ve got it and if you don’t, you don’t
 First compound of series dictates the trend.
28
Homing in on Residence Time in the Biaryl series
4-Azaindole
Me-4-Azaindole
GTPS IC50 (nM)
14
16
Dissociation t½ (h)
13
1.3
21
A Magic Methyl for SKR ?
29
Homing in on Residence Time in the Biaryl series
IC50 5 nM
IC50 8 nM
Dissociation
Di
i ti
half-life
9h
Dissociation
Di
i ti
half-life
11 h
HO
S
O
O
inactive
N
N
HN
IC50 4 nM
Dissociation
half-life
2 min
25
i
Long Residence Time in the Biaryls also comes from an H-Bond Acceptor in ortho
30
Amide Rotamers
50:50
by NMR
cis
IC50 14 nM
Dissociation
half-life 2 h
N
HO
O
N
HO
CF3
O
trans
OMe
O
CF3
O
OMe
inactive
IC50 27 nM
half-life
half
life 1.5 h
Long Residence Time in the Biaryls comes from an H-Bond Acceptor in a specific orientation
31
Can we check the H-Bond Acceptor location?
O
HO
O
N
F
3
Ab initio minimisation
32
Compound
1
2
3
GTPS IC50
20 nM
4 nM
6 nM
Dissociation t½
10 h
17h
1.7
18 h
N
Can we achieve truly long residence?
Position-by-position analysis of good
structural
t t l features
f t
for
f Residence
R id
Time
Ti
Minimum expression of
potency and long residence
Benzyl to bump
up lipophilicity
3º amide as
H-bond acceptor
Isoxazole as
H-bond acceptor
Acetic acid
O
HO
N
O
N
Chloro is
OK here
Cl
N
methyl stub
Azaindole
Azaindole
GTPS IC50 (nM)
14 nM
GTPS IC50 (nM)
Dissociation t½ (h)
1.5 h
Dissociation t½ (h)
2.5 nM
46 ± 15 h
cLogP
1.1
cLogP
4.8
cLogD
-1.1
cLogD
2.4
Once-a-day purely from Residence Time ?
33
CRTh2: Can Residence Time Help ?

For a p
purely
y antagonistic
g
effect,, prolonging
p
g g Receptor
p Occupancy
p
yp
prolongs
g PD effect

You will only really appreciate a PK-PD disconnection if Dissociation half-life >> PK terminal half-life

For GPCR antagonism,
g
, you
y should count on extending
g the PD effect byy half a half-life

We are pretty good at explaining what’s behind Structure-Activity Relationships (SAR)
Structure-Kinetic Relationships (SKR) are in their infancy and/or qualitative

To find Long Residence, you need to look for it
Efficacy
Residence
Time
34
Potency
Acknowledgements
Medicinal Chemistry
Juan Antonio Alonso
Mª Antonia Buil
Oscar Casado
Marcos Castillo
Jordi Castro
Paul Eastwood
Cristina Esteve
Manel Ferrer
Pilar Forns
Elena Gómez
Jacob González
GalChimia
Sara López
Marta Mir
Imma Moreno
Sara Sevilla
Bernat Vidal
Laura Vidal
Pere Vilaseca
ADME
Manel de Luca
Peter Eichhorn
Laura Estrella
Dolores Marin
Juan Navarro
M t Vives
Montse
Vi
Miriam Zanuy
35
Respiratory Therapeutic Area
Mó i Bravo
Mónica
B
Marta Calbet
José Luís Gómez
Encarna Jiménez
Martin Lehner
Isabel Pagan
Biological Reagents and Screening
Fani Alcaraz
Miriam Andrés
Julia Díaz
Teresa Domènech
Vicente García
Rosa López
Adela Orellana
Sílvia Petit
Israel Ramos
Enric Villanova
Villano a
Integrative Pharmacology
Clara Armengol
Laia Benavent
Luis Boix
Elena Calama
Amadeu Gavaldà
B t í Lerga
Beatríz
L
Sonia Sánchez
Computational and Structural
Ch i t and
Chemistry
d Bi
Biology
l
Sandra Almer
Yolanda Carranza
Sònia Espinosa
Estrella Lozoya
Pathology and Toxicology
Ana Andréa
Mariona Aulí
Cloti Hernández
Hernánde
Neus Prats
Development
Cesc Carrera
Nieves Crespo
Juan Pérez Andrés
Juan Pérez García
Gemma Roglan
Francisco Sánchez
Carme Serra
Intellectual Property
Houda Meliana
Eulàlia
u à a Pinyol
yo
Management
Paul Beswick
Jordi Gràcia
Vi t Matassa
Victor
M t
Monste Miralpeix

Download Report