physicochemical properties at bayer healthcare (wuppertal...
TRANSCRIPT
Jörg Keldenich Nov. 2006 2. PhysChem Forum 1
Physicochemical Properties at Bayer HealthCare (Wuppertal) and Their Use
in Medicinal Chemistry
Jörg Keldenich Nov. 2006 2. PhysChem Forum 2
Contents
Measured physico-chemical parameters
Introduction of our laboratory
Model systems for lipophilicity
Solubility
Use in medicinal chemistry
Jörg Keldenich Nov. 2006 2. PhysChem Forum 3
Physicochemical Properties Measured
Lipophilicity Membrane Affinity MA
Plasma binding human serum albumin binding HSArat serum albumin binding RSA
pKa pKa
Solubility screening in various buffers SOL
equilibrium in bufferequilibrium in galenic formulations
Jörg Keldenich Nov. 2006 2. PhysChem Forum 4
Logistics for HT Physicochemistry
BLJ input for sample identificationVial collecting rack
Bar-coded vial for sample registration
Jörg Keldenich Nov. 2006 2. PhysChem Forum 5
Data Handling by Laboratory Information and Management System (LIMS)
BAYNO/Prepno.,Barcode, Scale ofTests, Weight,Molecular Weight, Principleinvestigator, Projekt, Comparison
ChemistryLaboratories
PILO-LIMS
Lab.-Journal PIXPDH
LISSY Sample-preparation
LC/MS/MSWaters
Quattro-Micro
RoboterBar-code scan
Sample
Rackno.PositionBarcode
Methods
Rack
MTP
Sequences
SolubilityHSA-bindingMembrane affinity
Sample-registration
RackArea,Time
ReportsResults,Calibration Data
cMA, cHSAflag for
bases or acids
Archive
Jörg Keldenich Nov. 2006 2. PhysChem Forum 6
Our model system
Solid-supported lipid membranes (TRANSIL®)
Jörg Keldenich Nov. 2006 2. PhysChem Forum 7
Why Use Membrane Affinity?1. comparison with other lipophilicity descriptors
mlogP vs mlogMA
0.0
1.0
2.0
3.0
4.0
5.0
6.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
logMA
logP
mlogP (Österberg)
mlogP (Österberg) acid
mlogP (Österberg) basemlogD7.4 vs mlogMA
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
-4.0 -2.0 0.0 2.0 4.0 6.0
mlo
gD
mlogMA (Österberg) vs logMA Bayer
y = 0.9977x - 0.4014R2 = 0.8543
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0
logMA Bayer
logM
A Ö
ster
berg
et a
l.
Jörg Keldenich Nov. 2006 2. PhysChem Forum 8
Why use membrane affinity?2. comparison with physiological membranes
lipiophilicity vs MAerythrocytes
y = 0.1404x1.4994
R2 = 0.8551
y = 0.2121x1.3531
R2 = 0.7327
10
100
1000
10000
100000
1000000
10000000
100 1000 10000 100000
MAery
lipop
hilic
ity
MA measured
cPow
ACD D7.4
cMA
Influence of cholesterol content has to be considered
Jörg Keldenich Nov. 2006 2. PhysChem Forum 9
Why use membrane affinity?Influence of cholesterol
MA vs Molratio Lipid/Cholesterol
100
1000
10000
100000
1000000
0 0.2 0.4 0.6 0.8 1
Molration
MA
C. Tradum et al.:Biophysical J. 78 2496-2492
Jörg Keldenich Nov. 2006 2. PhysChem Forum 10
Why use membrane affinity?Influence of cholesterol
MA vs MA ERY
y = 0.1404x1.4994
R2 = 0.8551
0
100000
200000
300000
400000
500000
600000
0 5000 10000 15000 20000 25000
y = 0.6707xR2 = 0.9892
0
2000
4000
6000
8000
0 5000 10000 15000
MA measured with pure egg lecithin MA measured with cholesterol/egg lecithin ratio of 0.8
MA Ery MA Ery
Jörg Keldenich Nov. 2006 2. PhysChem Forum 11
Why use membrane affinity?Influence of cholesterol
Influence on passive permeation expected to be strong
flexibility of plasma membrane is strongly influence by cholesterol, content usually about 80 mol% of phospholipids content
Influence on distribution expected to be low
80% of all membranes are intracellular with a cholesterol content about 4 mol% of phospholipids content
Jörg Keldenich Nov. 2006 2. PhysChem Forum 12
Comparison of solubility methods
precipitation
• all compounds
• small amounts
• fast analytics
• compound dissolved in
organic solvent
• oversaturated solutions
possible
from powder
• selected compounds
• large amounts (two samples)
• specific analytics
• sensitive to morphology
• sensitive to purity
• sensitive to solvent
impurities
Jörg Keldenich Nov. 2006 2. PhysChem Forum 13
Comparison of solubility methods
precipitation
• Dissolve compound in DMSO
(2mg/40µl)
• Add 10µl of this solution to
1000µl buffer (1% DMSO)
• Shake for 24h at room
temperature
• Centrifuge to get supernatant
• Establish LC/MS/MS method
• Measure calibration standards
and probe
from powder
• Weight an appropriate amount
of compound as solid
• Add 1000µl buffer
• Shake for 24h at room
temperature
• Centrifuge to get supernatant
• Establish LC/MS/MS method
• Measure calibration standards
and probe
Jörg Keldenich Nov. 2006 2. PhysChem Forum 14
Comparison of solubility methods
EXAMPLES:
Compound SOL (precipitation) [mg/l] SOL (from powder) [mg/l]
Cpd 1 0.5 ± 0,2 0.5 ± 0,3Cpd 2 7.9 ± 1,2 1.3 ± 0.2Cpd 3 8.8 ± 3,8 4.2 ± 1.1Cpd 4 3.9 ± 0.7 0.6 ± 0.1
Cpd 5 mod Imod II
0.8 ± 0.081.5 ± 0.4
0.4 ± 0.081.1 ± 0.08
Cpd 6 mod Bamorphous
<0.1<0.1
<0.1<0.1
Cpd 7 mod Imod II
350 ± 18330 ± 26
420 ± 17380 ± 19
Jörg Keldenich Nov. 2006 2. PhysChem Forum 15
Comparison of solubility methodsName MOLSTRUCTURE
Yalkowsky(water)
charged neutral compounds
Warfarin 40 235
Primidone 500
Metolazone 60 88
Nifedipine 6 9
Ketoprofen 140 275
Glyburide 4 1.1
Indomethacin 8.6 240
Cimetidine 11000
Phenacetin 800 1080
Haloperidol 14 180
Phenytoin 26 23
Ibuprofen 36 290
Acetanilide 6300
OH
OOO
NHO
NH O
NH
NO
S
Cl
NH2
OO
N+
NH
OO
OOO
O
O
OHO
ONH
OCl
SNH
OO
NH
O
NO
OOH
O
Cl
NNH S NHN
NH
N
NH
OO
OH NCl O
F
NHNH
O
O
OHO
NH
O
Solubility [mg/l]: comparison PILO vs literatue (Yalkowsky)
1
10
100
1000
10000
1 10 100 1000 10000
literature (water)
PILO
(buf
fer 6
.5)
charged compoundsneutral compounds
Jörg Keldenich Nov. 2006 2. PhysChem Forum 16
Lessons learned from solubility comparisons
method differences not really critical
physical form very important
differences between research and development result from:
• morphology differences
• impurities
• solvent content
counter-ions and buffers are important when compound is
charged in solution
Jörg Keldenich Nov. 2006 2. PhysChem Forum 17
Case Histories: The Use of Physicochemical Properties
Two different projects as examples:
1. Reducing lipophilicity and HSA binding to increase fraction unbound: erectile disfunction
2. Influence of solubility on in vivo efficacy: the HSV project
Jörg Keldenich Nov. 2006 2. PhysChem Forum 18
Reducing Lipophilicity and Protein Binding to Increase Fraction Unbound
Starting point: initial compoundmoderate effective IC50 PDE-5: 530nM
DP1 compound: Vardenafilhighly effective IC50 PDE-5: 2nM
NH
NN
N
O
O
SN OO
N
NH
N N
N
O
Insufficient physicochemical properties:
high membrane affinity: 16500high protein binding: 1.7e-5 mol/lsolubility: below detection limitfraction unbound: <1%
no in vivo efficacy
improved physicochemical properties:
reduced membrane affinity: 580reduced protein binding: 1.2e-4 mol/lsolubility: 220 mg/lfraction unbound: 14%
excellent in vivo efficacy
Jörg Keldenich Nov. 2006 2. PhysChem Forum 19
Reducing lipophilicity and protein binding to increase fraction unbound
Even the prediction of the in vivo effect from IC50 and fractionunbound (calculated from MA and HSA) was possible
1.E-02
1.E-01
1.E+00
1.E+01
1.E+02
1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03Dose calculated from IC50 and physicochemical
properties
Dos
e m
easu
red
in v
ivo
initial compound
Vardenafil
Jörg Keldenich Nov. 2006 2. PhysChem Forum 20
Influence of solubility on in vivo efficacy
survival % of HSV infected mice at 60mg/kg vsfree serum normalized with IC50
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.00 0.01 0.10 1.00 10.00 100.00free serum/IC50
surv
ival
% a
t 60m
g/kg
SN
N
S
NH
O
N
O
O
NO
Starting point: Example 1moderate activity IC50: 750nM
Physicochemical properties:
Membrane affinity: 1430protein binding: 2e-4 mol/l
fraction unbound: 10%Solubility: 17mg/l
optimization of physicochemistry notnecessary, activity has to be improved
Jörg Keldenich Nov. 2006 2. PhysChem Forum 21
Influence of solubility on in vivo efficacy
S
N
NS
ONH
O
O
Example 2: brilliant compound in vitro IC50:
<1 nM
Physicochemical properties:
Solubility: <0.1 mg/l
excellent in vivo efficacy when
administered as solution, no in vivo
efficacy even as micronized powder
S
N
NS
ONH2O
ON
Development candidatein vitro activity IC50: 20 nM
Physicochemical properties:
Membrane affinity: 1590protein binding: 1e-5 mol/l
fraction unbound: 1%Solubility: 2.7mg/l
good in vivo efficacy
survival % of HSV infected mice at 60mg/kg vsfree serum normalized with IC50
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.00 0.01 0.10 1.00 10.00 100.00
free serum/IC50
surv
ival
% a
t 60m
g/kg
Jörg Keldenich Nov. 2006 2. PhysChem Forum 22
Conclusion
Impact of Physicochemistry Proven
Physicochemistry/ADME Implemented in Medicinal Chemistry
Properties Routinely Measured for Every Strategic Project
Use in Lead Optimization and Exploratory Research Established