Potent and selective, orally active GPBAR1 agonists as chemical biology probes

Potent and selective, orally active GPBAR1
agonists as chemical biology probes
Patrizio Mattei, F. Hoffmann-La Roche AG,
Pharma Research & Early Development
Acknowledgements
Discovery Chemistry
Kenneth Atz
Caterina Bissantz
Anne-Sophie Cuvier
Henrietta Dehmlow
Patrick Di Giorgio
Olivier Gavelle
Selina Hodel
Marie-Paule Imhoff
Kersten Klar
Michelle Kovacic
Rainer Martin
Patrizio Mattei
Ulrike Obst
Sherrie Pietranico
Noemi Raschetti
Flore Reggiani
Hans Richter
Martin Ritter
Marianne Rueher
Joana Salta
Petra Schmitz
Jitka Weber
Daniel Zimmerli
Cardiovascular/Metabolic DTA
Andreas Christ
Sabine Grüner
Maria Chiara Magnone
Tanja Minz
Anja Osterwald
Susanne Raab
Wolfgang Rapp
Astride Schnoebelen
Sabine Sewing
Urs Sprecher
Christoph Ullmer
Elisabeth Zirwes
Discovery Technologies
Thilo Enderle
Martin Graf
Arne Rufer
Non-clinical Safety
Rubén Alvarez Sánchez
Andreas Brink
Nicole Clemann
Jürgen Funk
Tobias Manigold
Axel Pähler
Formulation
Jochem Alsenz
Process Research & Synthesis
Stephan Bachmann
Manuela Baumann
Fritz Bliss
Christian Jenny
Jacky Joerger
Fritz Koch
Ernst Kupfer
Michel Lindemer
Karina Müller
Dieter Muri
Kurt Schönbächler
Markus Steiner
Literature
Werner Klaus
Patents
Dörte Klostermeyer
Management
Karin Conde-Knape
Silvia Pomposiello
Martin Stahl
René Wyler
Type 2 diabetes
• Type 2 diabetes
– a disease resulting from a combination of defective insulin secretion and
defective responsiveness to insulin.
– 285 M people (6.4% of world adult population) affected in 2010.
(source: International Diabetes Federation, http://www.diabetesatlas.org)
– a manageable chronic condition but risk of long-term complications.
• Initial therapy of type 2 diabetes starts with metformin (in addition, lifestyle changes
are strongly recommended: diet and exercise)
• If sufficient glucose control is not achieved within 3 months, add one of the following
to metformin (ADA/EASD consensus report: Diabetes Care 2012, 35, 1364):
– sulfonylurea (e. g., glibenclamide)
– PPARγ agonist (pioglitazone)
GLP–1 is a potent inducer of
– DPP4 inhibitor (e. g., sitagliptin)
glucose-dependent stimulation of
– GLP–1 receptor agonist (e. g., liraglutide)
glucose secretion
– insulin
Bile acids as incretin secretagogues
plasma PYY [pM]
PYY
plasma enteroglucagon [pM]
• Deoxycholic acid administered through intracolonic infusion in humans stimulates
PYY and GLP-1 release through activation of L-cells, to extent similar to that of a
normal meal (Gut 1993, 34, 1219).
enteroglucagon
• Discovery of the target, GPBAR1 (TGR5, M-BAR) independently at Takeda and Banyu (2002),
leading to intracellular increase of cAMP. Bile acids are endogenous ligands:
• deoxycholic acid EC50 = 1 µM
• lithocholic acid EC50 = 0.53 µM
• Bile acids promote GLP-1 secretion through GPBAR1 (BBRC 2005, 329, 386)
in STC1 enteroendocrine cell line
GPBAR1 action on L-cells and macrophages
Both anti-diabetic and anti-inflammatory effects
•
L-cells
Secretagogue:
PYY
GLP-1
•
•
Glucose-dependent glycemic
control
– low risk of hypoglycemia
– synergistic action of peptides
– parallels to bariatric surgery
Islet preservation
Body weight loss
cAMP
cytokine release:
TNFα
α
IL6
Kupffer cells
Monocytes
•
Anti-inflammatory actions
– Improved insulin sensitivity
– Reduced liver inflammation
– Reduced liver fibrosis
Reference compounds (1)
Synthetic bile acids
Intercept/Università di Perugia:
Obeticholic acid (INT-747)
O
chenodeoxycholic acid
core structure
OH
H
Intercept/Università di Perugia:
INT-777 [6-EMCA]
O
cholic acid
hydroxyl
OH
H
H
H
HO
H
OH
H
H
OH
hGPBAR EC50 0.5 µM
FXR: EC50 0.03 µM
• Mixed FXR/GPBAR1 agonist.
• Phase 3 (primary biliary cirrhosis);
Phase 2/3 (non-alcoholic steatohepatitis)
H
HO
H
H
OH
hGPBAR EC50 0.82 µM
FXR: inactive
• Selective GPBAR1 agonist.
• Anti-diabetic effects:
Cell Metabolism 2009, 10, 167.
• Reduction of atherosclerosis by reducing
macrophage inflammation and lipid
loading: Cell Metabolism 2011, 14, 747.
• "lead candidate to advance into clinical
studies" (J. Med. Chem. 2009, 52, 7958).
Reference compounds (2)
Non-bile acids
Exelixis/Bristol-Myers Squibb: XL-475
GlaxoSmithKline: SB-050
O
O
O
structure not disclosed
O
S N
O
O
N S
O
O
hGPBAR1 EC50 1.2 µM
• Orally administered GPBAR1 agonist. Designed
to selectively target the GPBAR1 receptors
in the intestine without significant systemic
exposure to enhance the therapeutic index for
potential chronic administration.
• Effective in lowering blood glucose, improving
glucose tolerance, improving plasma and
hepatic lipid levels, and reducing hepatic
steatosis.
•
•
•
Compound SB-050 progressed to the
clinic but was stopped due to inconsistent
PD effects/large inter-individual variations
(e.g. GLP-1 release) across doses.
Not clear whether systemic exposure is
required for PD effect.
241st ACS National Meeting & Exposition,
Anaheim 2011, MEDI-335.
Today's topics
• A new class of non-bile acid derived, orally bioavailable GPBAR1 agonists
Start from scratch, by way of high-throughput screening
• Characterisation of the new GPBAR1 agonists as antidiabetic agents
Glucose, PYY, GLP–1
GPBAR1 – HTS
• High-throughput screening of the whole Roche library (n = 940000)
• 5147 validated hits, of these 847 at hEC50 < 3 µM
• Examples (other than bile acids, steroids):
NK1 antagonists
(n = 43)
nicotinamides
(n = 1)
"Novartis-type"
(n = 4)
oximes
(n = 8)
O
Cl
F
N
F
N
S
O
N
O
Cl
O
F
F
O
N
O
N
N
N
F
N
OH
N
F
Cl
F3C
Cl
N
N
hEC50 [µM]
1.21
0.49
0.58
0.051
mEC50 [µM]
>20
2.7
>20
0.28
Ligand alignment: HTS hits/synthetic bile acid
H
HO
H
O
OH
O
H
Cl
H
O
H
N
S
O
O
F
O S N
O
O
hEC50 = 0.095 µM
J. Med. Chem.
2007, 50, 4266.
N
Cl
hEC50 = 0.48 µM
N
F
F
F
F
Cl
N
Cl
HBA
aromatic/hydrophobic:
essential feature
hydrophobic
N
hEC50 = 0.051 µM
N
hEC50 = 0.58 µM
aromatic
N
N
F
F
hEC50 = 1.21 µM
F
HO
O
N
N
O
aromatic/hydrophobic
anionic/polar
F
N
Should we consider oximes at all?
Marketed drugs
Hormonal contraceptives
Antibiotics
O
O
OH
O
O
S
H2N
N
O
N
N
H
H
H
H
S
N OH
HO
H
H
N
INN
cefdinir
norgestimate
marketed by
Astellas/Abbott + generics
Johnson & Johnson + generics
peak sales (year)
ca. USD 900 M (2006)
n/a
dose, administration route
oral, 300 -600 mg/day
Oral, 0.18 – 0.25 mg per tablet, predominantly
in combination with ethynylestradiol
half life
1.7 h
n/a
clogP
-0.5
5.1
metabolism
none (renal excretion of parent)
Complete first-pass metabolism (intestine,
liver) to active metabolites norelgestromine
(deacetylation), and norgestrel (deacetylation
+ ketone formation)
Resynthesis of the HTS hit
O
O
O
NaOH, RT (neat)
N
H
+
PhB(OH)2
Et2Zn
toluene, 60°C
N
Perkin 1 2001, 3258
N
NH2OH.HCl,
NaOAc,
EtOH/H2O
O
N
HO
N OH
+
N
N
E/Z ca. 6:1
N
N
N
N
N
N
H2N
HCl,
1,4-dioxane
O
+
N
Profile of the oxime hits
OH
N
N
N
human/mouse EC50 [nM]
N
OH
N
Br
45
19
2000
760
solubility [mg/L] (LYSA, pH 6.5)
<1
<1
logD (pH 7.4)
3.4
>3
aqueous stability (pH 1, 4, 6, 8, 10)
n. d.
stable
medium/high
medium/high
169 / 460
242 / 388
<0.2 / 3.1 / <0.2
<0.2 / 2.2 / <0.2
n.d.
5.5
mouse EC50 [nM]
permeability (prediction)
Clmic [mL/min/mg protein] (h/m)
CYPs [μM] (3A4, 2D6, 2C9)
Mouse PK
hERG IC20 [μM]
Cl [mL/min/kg]
77
Vss [L/kg]
1.6
F [%]
32
t1/2 [h]
0.2 – 0.5
protein binding fu [%] (h/m)
0.2 / 0.5
Initial SAR: Oxime, pyridine
X
N
N
OH
HO
N
O
N
O
N
0.045 µM
hEC50
0.26 µM
>10 µM
>10 µM
N OH
N
N
N
F
N
N
hEC50
0.045 µM
(–)-Enantiomer: 0.028 µM
(+)-Enantiomer: 0.105 µM
>10 µM
>10 µM
>10 µM
>10 µM
Optimisation of "northern" vector
N OH
N
N
hEC50 = 0.045 µM
N
hEC50
0.025 µM
0.077 µM
0.1 µM
0.38 µM
H
>10 µM
>10 µM
Lipophilic unsubstituted 4-pyridyl derivatives
3A4/2C9 pharmacophore
N OH
N
Br
hEC50
0.012 µM
CYP3A4, IC50
<0.2 µM
CYP2D6, IC50
4.3 µM
CYP2C9, IC50
<0.2 µM
logD
IC50 [µM]
CYP3A4 (n = 539)
>4
CYP2D6 (n = 520)
CYP2C9 (n = 497)
100
100
100
10
10
10
1
1
1
0.1
0.1
0.1
-2
-1
0
1
2
logD7.4
3
4
5
-2
-1
0
1
2
logD7.4
3
4
5
-2
-1
0
1
2
logD7.4
3
4
5
3-Methyl-4-pyridyl is not a CYP450 pharmacophore
R
N
IC50 [µM]
CYP3A4 (n = 554)
CYP2D6 (n = 540)
CYP2C9 (n = 538)
100
100
10
10
1
1
0.1
0.1
-2
-1
0
1
2
logD7.4
3
4
-2
5
logD7.4
-1
0
1
2
logD7.4
3
4
5
3-Methyl-4-pyridyl is not a CYP450 pharmacophore
hEC50
N OH
N
Br
IC50 [µM]
CYP3A4 (n = 554)
0.028 µM
CYP3A4, IC50
5 µM
CYP2D6, IC50
19 µM
CYP2C9, IC50
1.6 µM
logD
>4
CYP2D6 (n = 540)
CYP2C9 (n = 538)
100
100
10
10
1
1
0.1
0.1
-2
-1
0
1
2
logD7.4
3
4
-2
5
logD7.4
-1
0
1
2
logD7.4
3
4
5
Other head groups with reduced CYP interaction
N OH
R
Br
N
N
O
N
O
N
O
O
N
N
CONH2
COOH
hEC50
0.012 µM
0.028 µM
0.008 µM
>10 µM
0.179 µM
0.87 µM
CYP3A4, IC50
<0.2 µM
5 µM
>50 µM
15 µM
n. a.
>50 µM
CYP2D6, IC50
4.3 µM
19 µM
24 µM
25 µM
50 µM
>50 µM
CYP2C9, IC50
<0.2 µM
1.6 µM
5.1 µM
4.4 µM
32 µM
6.6 µM
>4
>4
3.8
3.8
2.9
1.0
logD
Introduction of polarity at "south eastern" exit vector
N OH
N
COOH
Br
COOH
COOH
H
N
COOH
O
hEC50 = 0.028 µM
0.26 µM
0.7 µM
0.02 µM
0.29 µM
0.01 µM
mEC50 = 0.74 µM
3.7 µM
6.1 µM
0.093 µM
0.86 µM
0.07 µM
1.1
0.9
2.5
2.4
1.0
medium/high
medium/high
medium/high
medium/high
LOW
logD = >4
permeatility = medium/high
N OH
N OH
N
N
COOH
hEC50
0.011 µM
COOH
0.057 µM
COOH
Comparison of in vitro profiles
HTS hit
benzoic acid lead compound
O
OH
N
OH
N
Br
OH
N
N
human EC50 [nM]
19
11
mouse EC50 [nM]
760
130
inactive
inactive
off-target activity: PPARs, LXR, PXR, RXR
n.d.
inactive
solubility [mg/L] (LYSA, pH 6.5)
<1
9
logD (pH 7.4)
>3
2.7
stable
stable
medium/high
medium/high
n.d.
16 / 53
<0.2 / 2.2 / <0.2
45 / >50 / 14
5.5
>10
0.2 / 0.5
0.04 / 0.2
FXR transactivation
aqueous stability (pH 1, 4, 6, 8, 10)
permeability (prediction)
Clhep [mL/min/mg protein] (h/m)
CYPs [μM] (3A4, 2D6, 2C9)
hERG IC20 [μM]
protein binding fu [%] (h/m)
Improved properties translate into an improved PK
HTS hit
benzoic acid lead compound
O
OH
N
OH
N
Br
100000
10000
plasma conc. (ng/mL)
plasma conc. (ng/mL)
N
p.o. 8.1 mg/kg
i.v. 2.2 mg/kg
1000
100
10
OH
N
10000
p.o. 23 mg/kg
i.v. 1.0 mg/kg
1000
100
10
1
1
0
2
4
6
Time (h)
Cl
Vss
t1/2
[mL/min/kg]
[L/kg]
[h]
77
1.6
F
8
cmax norm.
[%] [ng/mL]/[mg/kg]
0.2-0.5 32
0
65
mouse
2
4
6
Time (h)
8
Cl
Vss
t1/2
F
cmax norm.
[mL/min/kg]
[L/kg]
[h]
[%]
[ng/mL]/[mg/kg]
15
0.5
2.5
80
800
PK/PD for benzoic acid lead compound
PYY as mechanistic readout
O
OH
N
C57BL6 mice
OH
N
PK/PD relationship
concentration vs. time
dose: 100 mg/kg p.o.
3000
concentration
exposure [ng/mL]
PYY [pg/mL]
10000
1000
Total PYY [pg/mL]
100000
2500
p.o. 100 mg/kg
p.o. 50 mg/kg
i.v. 2 mg/kg
2000
1500
1000
500
100
0
5
10
15
time [h]
20
25
0
10
100
1000
10000
exposure [ng/mL]
• No indirect or delayed effect
• Plasma exposure appears to be a good surrogate for GPBAR1 action.
100000
Oral glucose tolerance test of lead compound
Dose dependent improvement of glucose tolerance
§§ p<0.01 Jonckheere's trend test supports
10
0
O
OH
N
-10
-10.4%
OH
N
-20
-18.9%
-22.8%
§§
∆Glucose AUC0-120min [%]
+5%
C57BL6 mice
high probability of dose dependent effects
-30
20
50
70
Dose [mg/kg]
100
compound administration 1 h
before glucose challenge
Second generation: Reduce logD/protein binding
hEC50
in vitro:
O
OH
N
mEC50 solubility logD fu (h/m)
[nM]
[nM]
11
Cl
OH
mouse PK:
[mg/L]
[%]
130
Vss
9
t1/2
F
2.7 0.04/0.2
cmax norm.
[mL/min/kg]
[L/kg]
[h]
[%]
[ng/mL]/[mg/kg]
15
0.5
2.5
80
800
N
O
O
OH
N
OH
N
N
OH
N
OH
N
OH
S
O
O
N
N
O
O
RO5527239
hEC50
mEC50 solubility logD fu (h/m)
[nM]
4
Cl
[nM]
[mg/L]
[%]
28
Vss
273
1.5 0.4/1.9
cmax norm.
t1/2
F
[mL/min/kg]
[L/kg]
[h]
[%]
[ng/mL]/[mg/kg]
5
0.4
1.7
100
650
hEC50
mEC50 solubility logD fu (h/m)
[nM]
60
Cl
[nM]
[mg/L]
[%]
450
Vss
250
t1/2
F
1.3 0.4/0.4
cmax norm.
[mL/min/kg]
[L/kg]
[h]
[%]
[ng/mL]/[mg/kg]
5
0.5
0.7
87
600
hEC50
mEC50 solubility logD fu (h/m)
[nM]
26
Cl
[nM]
[mg/L]
[%]
1300
Vss
300
1.7
2.1/11
cmax norm.
t1/2
F
[mL/min/kg]
[L/kg]
[h]
[%]
[ng/mL]/[mg/kg]
67
3.3
2.2
29
40
Synthesis of RO5527239
CO2Et
NC
O
ClMg
Morpholine,
∆, 2h
EtO
H
Br
O
O
15 h, 110°C
EtO
CN
87%
CH3CO2H, H2SO4
O
HO
CN
2-Me-THF,
RT, 1h
quant.
Br
72%
Br
Br
Br
O
HCl
n-Heptane/i-PrOH 80:20
TBTU, NEt3, DMF
91%
N
OMe
(rac)
n-BuLi,
2-Me-THF,
-100°C
N
Reprosil Chiral NR
HNMe(OMe)
O
N
OMe
Br
O
N
73%
Br
Br
46% (R)
42% (S)
1. HN
COOEt
1. NH2OH.HCl, NaOAc,
EtOH, reflux
2. HCl, 1,4-dioxane
O
Pd2(dba)3, X-Phos
NaOtBu, toluene
2. NaOH, EtOH, H2O
76%
N
N
N
OH
N
N
76%
O
O
OH
OH
RO5527239
RO5527239 binds specifically to GPBAR1
Same binding site as lithocholic acid
O
OH
N
T
OH
No meaningful off-target activity identified
[3H]-RO5527239
N
specific binding [pmol/mg]
N
7
6
5
4
3
Kd = 70 nM
2
Bmax = 10.2 pmol/mg protein
1
0
0
20
40
60
80
free radioligand (nM)
100
•
radioligand binding panel (Cerep, n = 97)
•
transactivation assays:
•
FXR
•
LXRα, LXRβ
•
PPARα, PPARγ, PPARδ
•
RXRα
Cellular activity of RO5527239
GPBAR1 expressing enteroendocrine STC-1 cells
cAMP
GLP–1
O
OH
N
EC50 = 1.59 µM
N
EC50 > 10 µM
OH
EC50 = 0.32 µM
RO5527239
EC50 = 11 µM
deoxycholic acid
GLP-1 (% of PMA)
cAMP (nM)
160
N
120
80
40
0
0.1
1
10
concentration (µM)
28
24
20
16
12
8
4
0
-4
0.1
1
10
concentration (µM)
RO5527239 in humanised DIO (C57/Bl6) mice
GLP-1/PYY secretion and glucose tolerance
benzoic acid lead compound
O
OH
N
N
OH
10 mg/kg
30 mg/kg
N
100 mg/kg
N
100
total GLP-1 [pg/ml]
total PYY [pg/ml]
800
400
80
60
40
20
60 min
OH
10
Vehicle
Glucose
20
-31 -28 -37%
8
6
4
2
0
18
16
14
12
10
8
6
0
0
∆ AUC Glucose
GLP–1
PYY
1200
O
Plasma Glucose [mM]
OH
N
[mM×min]×100
RO5527239
60 min
4
0
30
60
min
• Higher in vivo efficacy of RO5527239 in GLP-1/PYY release translates into
superior glucose tolerance.
120
Acute improvement in glucose tolerance in db/db mice
Synergistic effects with metformin and sitagliptin
RO5527239 (30 mg/kg)
metformin (100 mg/kg)
glucose [mM]
vehicle
20
metformin (100 mg/kg)
+ RO5527239 (30 mg/kg)
16
0
8
N
OH
RO5527239
50
1
N
40
30
20
0
60
time [min]
120
−
−
25
+
−
+
+
−
+
RO5527239
metformin
Active GLP-1
20
30
15
fold increase
sitagliptin (10 mg/kg)
+ RO5527239 (30 mg/kg)
Total GLP-1
4
glucose [mM]
sitagliptin (10 mg/kg)
OH
N
10
b) with sitagliptin
(DPP4 inhibitor)
RO5527239 (30 mg/kg)
-24%
2
12
-120 0 15 30
vehicle
O
fold increase
a) with metformin
[mM×min] × 1000
AUC Glucose
3
10
5
0
25
20
15
10
5
0
-60
0 15 30 60
time [min]
120
−
−
+
−
+
+
−
+
RO5527239
sitagliptin
Summary
• We have identified a new class of orally available GPBAR1 agonists starting
from high-throughput screening hits.
• An unsubstituted oxime group is required for this class of compounds.
• Optimised GPBAR1 agonists bind to the target in the same manner as bile
acids, resulting in secretion of PYY and GLP-1.
• RO5527239, the most potent compound of this class, improves glucose
tolerance in rodents and shows synergistic effects with sitagliptin and
metformin.
We Innovate Healthcare

Download Report