biophysical models for prediction of activity and …mbermejo/hawaii.pdfbiophysical models for...

1

Molecular Biopharmaceutics. 2003 Hawaii Biophysical models-ADME predictions

1© Marival BermejoDepartamento Farmaciay Tecnología Farmacéutica

Biophysical Models for Prediction of Activity and

ADME Parameters: Application to Quinolone

Antibiotics

Mª del Val Bermejo Sanz, Ph.D.Associate Professor of Pharmaceutics

Depto. Farmacia y Tecnología FarmacéuticaFacultad de Farmacia. Universidad de Valencia

España



OutlineBiophysical absorption models

QSAR methods in drug absorption and activity

Oral absorption prediction in drug development

C1

C2

C3

C4C6

C5C7

N N

O

COOHF

N

CH2

S N

O N

CH30.00

1.82

0.27

0.54

F0 : 0.96

-2.56

0.00

0.81

0.54

C0

*C1 : 0.00

0.54

0.54

0.54

0.54

Time (min)

0 100 200 300 400 500

Conc

entr

atio

n (µ

g/m

L)

0.1

1

10

100

BolusOral

Plasma

Lumen

2



Closed Loop Perfusion Technique: Doluisio’s method

Cannula

Stopcock

Syringe

Gut

SamplesDrug in SOLUTION•Bile duct closed•Water reabsorption measurement

Absorption rate coefficient:

0cka tA A e− ⋅= ⋅

2a

effk RP ⋅

=

Permeability values



Closed Loop Perfusion Technique: Doluisio’s methodAbsorption rate constant:

0cka tA A e− ⋅= ⋅

2a

effk RP ⋅

=

Permeability values

5 10 15 20 25 300.15

0.20

0.25

0.30

0.35

0.40

Time (min-1)

Con

cent

ratio

n(µ

g/m

L)

3



Biophysical Absorption Models: Plá-Delfina and Moreno

ka1

membrane

ka

Aqueous layer

lumen

Plasma

ka2aqueous pathway

ka

lipophilicity

da

a

PCCkp

PBPkmKa

+⋅+

+⋅=

Small Intestine Colon

membrane

ka

Aqueous layer

lumen

Plasma

Lipophilicity

ka

a

a

PBPkmKa

+⋅=



n-Octanol Partition coeffcient

0.01 0.1 1 10 100 1000 10000

Abso

rptio

n ra

te c

oeffi

ecie

nt h

-1

0

2

4

6

8

10

12Carbamates

Chromatographic constants

0.1 1 10 100

Abs

orpt

ion

rate

coe

ffici

ent h

-1

0

1

2

3

4

5Phenylalkylamines

n-Octanol Partition Coefficient

0.1 1 10 100 1000 10000 100000

Abs

orpt

ion

rate

coe

ffici

ent h

-1

0

2

4

6

8

10

Ring-Anilines

Chromatographic constant

0.1 1 10

Abs

orpt

ion

rate

coe

ffici

ent h

-1

0

1

2

3

4

5Sulfonamides MW>250 Da

Small intestine

4



HPLC, Capacity factor K'

0.1 1 10 100

Abso

rptio

n ra

te c

oeffi

cien

t h-1

0.50.6

56

1

Distribution coefficient

1 10 100 1000 10000

Abso

rptio

n ra

te c

oeffi

cien

t h-1

0

1

2

3

4

5

Colon

Ring-AnilinesPhenyl-alkyl acids



10 100 1000 10000

1

Distribution Coefficient, DpH:3.00

Gast

ric

abso

rpti

onra

teco

nsta

nt, h

-1

2

0,5

(CH2)n

n=1-7

COOH


5



Lumenaqueouslayer

MembraneGlycocalyxPhospholipids

Plasma ka

P

f

d

a PEPCk

1 ⋅+⋅=

Gastric mucosa




Na TAUROCHOLATE: INTERPRETATION OF ITS ULCEROGENIC EFFECT.

Distribution Coefficient (n-Octanol)10 100 1000 10000G

astr

ic a

bsor

ptio

n co

nsta

nt h

-1

1

kako

Lumen aqueouslayer

MembraneGlycocalyx

Plasma

Lumen aqueouslayer

MembraneGlycocalyxPhospholipids

Plasma

6



Biophysical Absorption Models: Summary

ka1

ka2

membrane

ka

Aqueouslayer

lumen

Plasma

ka

P

da

a

PCCkp

PBPkmKa

+⋅+

+⋅=

Small Intestine

membrane

ka

Aqueouslayer

lumen

Plasma

P

ka

a

a

PBPkmKa

+⋅=

Colon

Membrane

PlasmaLumen aqueous layer

GlycocalyxPhospholipids

Stomach

ka

P

f

d

a PEPCk

1 ⋅+⋅=



Biophysical Absorption Models: Summary

membrane

ka

Aqueouslayer

lumen

Plasma

P

ka

a

a

PBPkmKa

+⋅=

Colon

Membrane

PlasmaLumen aqueous layer

GlycocalyxPhospholipids

Stomach

ka

P

f

d

a PEPCk

1 ⋅+⋅=

ka

ka1

Pa

akm PKaB P

⋅=+

Small Intestine: MW> 250 Da

membrane

ka

Aqueouslayer

lumen

Plasma

7



F

N

O

COOH

N

NR

F

N

O

COOH

N

NR

R= H, Norfloxacin R= H, Ciprofloxacin

R= CH3 , CH3-(CH2 )n -

n=1-5CENAVISA S.A. Spain




0.01 0.1 1 10 100 1000 100000

1

2

3

4

5

6

7

0.01 0.1 1 10 100 1000 100000

1

2

3

4

5

6

7

Norfloxacin Ciprofloxacin

Partition coefficient, PPartition coefficient, P

Abs

orpt

ion

rate

cons

tant

h-1

Abs

orpt

ion

rate

cons

tant

h-1


8



Kaq (h-1) Kmem (h-1) 7.073 0.195 6.944 0.397 6.923 3.322 6.801 4.070 6.782 4.795 6.667 6.210 6.649 14.813 6.541 15.359 6.524 30.267 6.422 33.761 6.405 79.701 6.309 85.387 6.293 181.075 6.201 233.991 6.186 389.164 6.099 545.016

0.01 0.1 1 10 100 1000 100000.01

0.1

1

10

100

1000

Abs

orpt

ion

rate

cons

tant

h-1

Partition coefficient, P

kaq

kmem

kmemkaqka111 += 1

da d

C PkE MW P

⋅=+ ⋅ ⋅

membrane

K

Aq. layer

lumen

Plasma



F

N

O

COOH

OCF

N

O

O H

N

S

CENAVISA 8804 CENAVISA 8919

Flumequine

QSAR methodologies in drug absorption

OCF

N

O

O H

N

O

9



ITODYS. Université Paris VII. France. Prof. Christiane Mercier

n-Octanol Partition Coefficient

1x10-3 10x10-3 100x10-3 1x100 10x100 100x100 1x103 10x103 100x103 1x106

Pef

f (c

m/s

)

0.00

5.00e-5

1.00e-4

1.50e-4

2.00e-4

Flumequine

CNV 8804CNV 8919

Norfloxacin derivatives

Ciprofloxacin derivatives




CH3-COOH

CH3-CH2-CH2-CH3

CH3-CH-CH3

CH3

1

2

3

FO-A1-B11-C1

A2 B12Traze

Vector A1 B11 C1B12

0 1 0 0 0

A2

1 0 1 0 1

1 0 1 1 0

Vizet P. PATQSAR©. Version 2.55. 1997

ITODYS. Université Paris VII: Vizet P. PATQSAR©. Version 2.55. 1997


10



000000.02157140.8154290.837NEW001

00010-0.02278570.8747860.852NEW004

000100.001214290.8747860.876NEW003

001111.39E-160.0920.092NEW002

31011-0.01787210.8018720.784BER016

31011-0.02187210.8018720.78BER015

31011-0.02087210.8018720.781BER014

31011-0.04687210.8018720.755BER013

210110.01960370.6613960.681BER012

110110.02607950.5209210.547BER011

010110.02555530.3804450.406BER010

000110.0578214-0.259821-0.202BER009

310010.0354850.7425150.778BER008

310010.0344850.7425150.777BER007

310010.0124850.7425150.755BER006

310010.005485020.7425150.748BER005

210010.007960830.6020390.61BER004

11001-0.02256340.4615630.439BER003

01001-0.03708760.3210880.284BER002

00001-0.0578214-0.319179-0.377BER001

EQ(V6;V7;V8)V5V4V3V2Log Ka diff

Log Ka calc

Log Ka expCompounds

Topological matrixDependent variable



C1

C2

C3

C4C6

C5C7

N N

O

COOHF

N

CH2

S N

O N

CH30.35

0.64

0.140.140.00

0.00 0.00 0.14

F0 : 0.82

-1.13

0.00

0.00

0.06

C0

C1

C2

C3

C4C6

C5C7

N N

O

COOHF

N

CH2

S N

O N

CH30.00

1.82

0.27

0.54

F0 : 0.96

-2.56

0.00

0.81

0.54

C0

*C1 : 0.00

0.54

0.54

0.54

0.54

Lipophilicity-structure correlation


Absorption-structure correlation

Log P

Log ka

11



C1

C2

C3

C4

C5

C6

C7Fo : 1.03

0.00

CH2

N

F

NN

OCOOH

-0.28 -0.28

-0.28

-0.28

0.00 -0.28

0.56

C1

C2

C3

C4

C5

C6

C7Fo : 0.33

0.00

CH2

N

F

NN

OCOOH

-0.28

-0.28

-0.28

-0.28

-0.28

0.28

0.00

Structure- antibacterial activity correlation


E.Coli

Staph.

Log 1/CMI



Topoisomerase

Quinolone

12



Predicted Log Ka-0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0

Exp

erim

enta

l Log

Ka

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

R2=0.99

Predicted Log P-2 -1 0 1 2 3 4

Exp

erim

enta

l Log

P

-2

-1

0

1

2

3

4

R2=0.997

Predicted Log 1/Mic E. Coli-1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Expe

rimen

tal L

og 1

/Mic

E. C

oli

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

R2=0.96

Predicted Log 1/Mic Staph.-1.0 -0.5 0.0 0.5 1.0

Exp

erim

enta

l Log

1/M

ic S

taph

.-1.4-1.2-1.0-0.8-0.6-0.4-0.20.00.20.40.60.8

R2=0.95



1. Lipophilicity is the main factor governing quinolone absorption.

2. The absorption-partition relationship has a good predictive performance of intestinal permeability.

3. The antibacterial activity is decreased by the N’ alkyl chain which hinders the access to the bacterial topoisomerase or drug binding to DNA.

13



F

N

O

COOH

N

NR

F

N

O

COOH

N

NR

R= H, Norfloxacin

R= H, Ciprofloxacin

R= CH3 , CH3-(CH2 )n - n=1-5

Homologous CompoundsCENAVISA S.A. Spain

N

F

NN

CH3

R

COOHO

CNV 97104-(CH2)3-CH3

CNV 97103-(CH2)2-CH3

CNV 97102-CH2-CH3

CNV 97101-CH3

CNV 97100H

NameR



n-Octanol Partition coefficient10-2 10-1 100 101 102 103 104 105

P eff

(cm

/s)

0.0

5.0e-5

1.0e-4

1.5e-4

CNV9710X derivatives

CNV97100

CNV97103

CNV97104

CNV97102

CNV97101

14



n-Octanol Partition coefficient10-2 10-1 100 101 102 103 104 105

P eff

(cm

/s)

0.0

5.0e-5

1.0e-4

1.5e-4

CNV9710X derivatives

CNV97100

CNV97103

CNV97104

CNV97102

CNV97101



0

5e-6

1e-5

2e-5

2e-5

3e-5

3e-5

500 50

5 0.5

0.05

DuodenumJejunumIleumColonWhole Intest.

Per

mea

bilit

y cm

/s

CNV97100 conc. µg/mL

P-gp levels

15



0.0

5.0e-6

1.0e-5

1.5e-5

2.0e-5

2.5e-5

3.0e-5

3.5e-5

500 50

5 0.5

0.05

DuodenumJejunumIleumColonWhole Intest.

Per

mea

bilit

y cm

/s




0.0

5.0e-6

1.0e-5

1.5e-5

2.0e-5

2.5e-5

3.0e-5

3.5e-5

500 50

5 0.5

0.05

D J IleumWhole Intest.

Per

mea

bilit

y cm

/s


16



0.0

5.0e-6

1.0e-5

1.5e-5

2.0e-5

2.5e-5

3.0e-5

3.5e-5

500 50

5 0.5

0.05 D

JIleum

CWhole Intest.

+Verapamil

Per

mea

bilit

y cm

/s




Model assumptions

Pdiff unchanged along GI tract.Efflux system Km value (affinity) is unchanged along GI tract.Vm= f(expression level).

Ileum

17



max

max

Whole intestine2 2( )

Ileum2 2( )

diffm

d

a

a

i f

a

fb

b b m

CdC P Cdt K C

CdC P Cd

Vt KR C

VR

R

R−

−

⋅= − ⋅ ⋅ + ⋅+

⋅= − ⋅ ⋅ + ⋅+

Time (min)0 5 10 15 20 25 30 35

Con

cent

ratio

n (µ

g/m

l)

0.01

0.1

1

10

100

1000

3.06·10 - 4 (1.76·10 - 4)Vm (nmol·s-1·cm-2)Whole intestine

15.94 (8.25)Km (µM)

3.94·10 - 4 (2.06·10 - 4)Vm (nmol·s-1·cm-2)ileum

3.07·10 - 5 (0.20·10 - 5)Pdiff(cm·s-1)

In situ studyValue (SD)

Parameter



3. CNV97100 shows a saturable efflux process verapamil sensitive.

4. The quinolone effective permeability is lower in rat ileum due to the higher P-glycoprotein expression level.

18



)()( hg EEFaF −⋅−⋅= 11Eg= gut first-pass effects

Eh= liver first-pass effects

Oral absorption prediction in drug development“Bioavailability prediction in drug development: fluoroquinolones. CICYT

(SAF 96-1710)” Director. Prof. Plá-Delfina



)()( hg EEFaF −⋅−⋅= 11


1 ka TaF e− ⋅= −

)2(1

ef resf Ta

PRF e

− ⋅= −

( 2 )1

r

bs

es

aTT

aF e− ⋅

= −

(2 ) 1 nAaF e− ⋅= −

19



jugularVein

Blood Samples

Administration(I.V.)

Cannula

Oral Administration




Oral administration

Group A (8-12 rats)

I.V administration

(bolus or perfusion)

Group B (8-12 rats)

Plasma sampling and

Analysis

AUC0∞

calculation

AUC0t + AUCt

∞

(Trapezoidal rule) ( Ct/k)


PK Population analysis: Nonmem v.5.0Post-hoc Bayesian ind. parameter estimation

20



Time (min)0 100 200 300 400

Nor

floxa

cin

plas

ma

leve

ls (m

g/L)

0.01

0.1

1

10

100Div=2 mgDoral=4 mg

(1)

Time (min)0 100 200 300 400

Cip

roflo

xaci

n pl

asm

a le

vels

(m

g/L)

0.01

0.1

1

10

100DiV=4 mgDoral=4 mg



Time (min)0 100 200 300 4004N

'-Pr

opyl

-cip

roflo

xaci

n pl

asm

a le

vels

(mg/

L)

0.01

0.1

1

10

100 DiV=4 mgDoral=4 mg

(3)

Time (min)0 100 200 300 400 500 600

Peflo

xaci

npl

asm

a le

vels

(mg/

L)

0.01

0.1

1

10

100DiV=8 mgDoral=8 mg

(5)

21



ka in situ

0 1 2 3 4 5 6 7 8

F in

viv

o

0.2

0.4

0.6

0.8

1.0

1.2

Norfloxacin

Ciprofloxacin

CNV97100

Pefloxacin

CNV97102

Propil-NorPropil-Cipro

CNV97104

Flumequine

CNV97103

CNV97101

1.041 kaaF e− ⋅= −



Predicted F 0.0 0.2 0.4 0.6 0.8 1.0 1.2

Exp

erim

enta

l F

0.0

0.2

0.4

0.6

0.8

1.0

1.2

b[0]=0.022b[1]=0.936r ²=0.927

1.041 kaaF e− ⋅= −

22



Cldis=k12·VcR2 = 0.6741

00.20.40.60.8

11.21.41.6

0.01 0.1 1 10 100

Cl=k10·VcR2 = 0.2145

00.10.20.30.40.50.60.70.80.9

1

0.01 0.1 1 10 100

Vc R2 = 0.0505

00.10.20.30.40.50.60.70.80.9

1

0.01 0.1 1 10 100

Vp R2 = 0.0082

00.20.40.60.8

11.21.41.6

0.01 0.1 1 10 100



4. The in situ intestinal permeability is a good predictive parameter of in vivo Fa.

5. The saturable (efflux and absorption) processes observed in situ (and in vitro) had not a significant influence in vivo in rats at the oral doses used.

6. Disposition parameters……………still under modelling

23



ATP

ADP+P

Passivedifussion

Activetransport

Transporter

ATP consumeVm ap; Km

SAz

ATP synthesis

ATPADP+P

ATP

F0F1 complex

Electron transport complexes

ATP levelsKinKout·[ATP]

Kd

SAz

Molecular mechanisms

Model construction

0.0

0.2

0.4

0.6

0.8

1.0

5025

51

0.5 0.03 mM0.3 mM

3 mM

Ka (h-1)

Sarafloxacin Conc Time [min]0 20 40 60

ATP

frac

tion

0.0

0.2

0.4

0.6

0.8

1.0

predictions

Data fitting

[ ] [ ]in out

d ATPk k ATP

dt= − ⋅

[ ] [ ] 0in out

d ATPk k ATP

dt= − ⋅ =

[ ] 1 in outATP k k= → =

[ ][ ]50

1inap in

SAzk k

C SAz

= ⋅ − +



Acknowledgments

Funds:− Generalitat Valenciana (GV 99-99-1-12)− MEC (SAF 96 1710)

Compounds: Dr. Joan Freixas CENAVISA SA

Isabel GonzálezCarlos Fernández

Vicente CasabóVirginia MerinoAna Ruiz

Ana

Isabel Vicente

Virginia

Carlos

biophysical models for prediction of activity and …mbermejo/hawaii.pdfbiophysical models for...

Documents