RSC, February 2014
Interplay between enzymes and
transporters in defining hepatic drug
clearance and intracellular
concentration of drugs
J Brian Houston
Centre for Applied Pharmacokinetic Research
(CAPkR)
Three elements of mechanism-based
prediction of human PK
In vitro kinetics in
multiple systems
ADME & DDI
Predictions
(static and
dynamic)
Modelling &
Simulation
(PBPK)
Generic view of drug kinetics in hepatocytes:
Various processes defining drug
intracellular concentration
Intracellular binding & trafficking
Efffux
Transporters
•
•
Challenges
To describe all processes
To understand
interplay & hence holistically
Media
model
D
Passive
diffusion
D
D
Intracellular
unbound conc
M
Clearance by microsomal/cytosolic enzymes
Extending Classic Hepatic Clearance Models:
Use of CLint,app to delineate transporters &
enzymes and their ‘Interplay’
Prediction (static) equations well established:
e.g. well-stirred
ll ti d liver
li
model
d l
Qh  fub  CLint
CL 
Qh  fub  CLint
Extended with use of Clint,app
p hepatocellular
p
sequential
q
pprocesses
int app to encompass
(Interplay model) based on Sugiyama and Pang.
CLint,app  CLint,met
CLint,active  CLint,pass
CLint,met  CLint,pass  CLint,eff
CLint,app
int app – Interplay of transporters and
enzymes
 High passi
passive
e permeability
permeabilit
 reduces to CLint,met
 Low passive permeability (with minimal effux)
 reduces to CLint,active
 Various intermediate cases. The second term collective –
Kpu (partition coefficient for unbound drug)
In vitro tools for assessment of
hepatic uptake
• Hepatocytes:
 Suspension culture - direct cell uptake (oil separation)
 Plated cells (also sandwich configuration)
 Single
Si l time
i
points
i
or ffullll time
i
course off uptake
k & metabolism
b li
• Comparative scaled activity in hepatocytes relative to
microsomes
i
( b ll l preparation)
(subcellular
ti )
• Sometimes rat better option than human
 Higher activity and less confounding issues surrounding
preparation and storage
 Minimal inter-individual variation
 Potential extrapolation to human
• CL terms main metric,, and for inhibition DDIs Ki
where CLi = CLcontrol / 1+Ki
Characterisation of extent of hepatic uptake
What are we measuring with Kpu?
•
Kpu = Cell to medium (plasma) unbound concentration ratio
Kpu
Kp
pu 

CLint,uptake  CLint, passive
CLint, passive
[True Kpu - at steady state when no
metabolism or efflux]
CLint, pass  CLint,uptake
CLint, pass  CLint, efflux  CLint, met
[Apparent Kpu]
• Contrasts with Kptotal (ratio of total concentrations)
which reflects both uptake and intracellular binding
K u = fu
Kp
f cell · Kp
K total
Used together estimates intracellular drug concentration
Comparison of microsomes and hepatocytes
E id
Evidence
ffor h
hepatic
ti uptake
t k affecting
ff ti
CL and DDI prediction
If active uptake process occurring, substrate or inhibitor
may show higher ‘affinity’
affinity in hepatocytes compared to
microsomes –
i e lower Km or Ki as
i.e.
[S]u,plasma << [S]u,liver or [I]u,plasma << [I]u,liver
K p ,u
Ki ,microsome

Ki ,hepatocyte
p
y
Km,microsome

Km,hepatocyte
CLint,
microsome
m
cr m
 nt,hepatocyte
CLint,hepatocyte
microsome
Impact of hepatic intracellular binding?:
Ki Microsomes vs. Hepatocytes
Ki Hepa
atocytes (µM)
100
((n= 7 inhibitors,, n=21 ppathways)
y)
Inhibitor
Cell-to-Media
Ratio
(Kptotal)
MCZ
FXT
KCZ
FVX
QUI
OMP
FCZ
6000
2010
1200
577
143
16
4.2
10
1
0.1
0.01
0.01
0.1
1
10
Ki Microsomes (µM)
 Miconazole
Fluconazole
Ketoconazole
Fluoxetine
Fl
 Quinine
Fluvoxamine
Fl
 Omeprazole
Saquinavir
Nelfinavir
Ritonavir
100
 Good agreement between Ki
values in both systems (both
corrected for non specific binding)
 Kp
K u approximately
i t l 1
Brown et al, DMD 2007
Lack of impact
p
of hepatic
p
uptake:
p
Micro
osome Ki / Hepato
ocyte Ki
Microsomal & Hepatocyte Ki ratio vs. Kptotal
4
3
2
1
0
10
100
1000
10000
Kp
C/M
t t l Ratio
total
 Kptotal differ over 3 orders of magnitude
 No correlation between Ki ratio and Kp
ptotal ((n = 27))
Brown et al, DMD 2007
Pr
a
R vas
o s ta
uv tin
At ast
or ati
n
v
C ast
er at
iv
in
Pi asta
ta
va tin
Sa sta
qu tin
in
R avir
i
Te ton
lm avi
r
i
O sar
lm ta
es n
ar
t
V
al an
C
la
s
rit art
h
a
Er rom n
yt
hr ycin
o
R my
ep
c
ag in
N lin
a
id
Fe teg e
xo lini
fe de
na
Bo din
se e
nt
an
Kpu ((CLuptake/P
Pdiff)
Kpu = 1 – importance of uptake transporters
for 16 drugs in rat (Yabe et al DMD 2011)
10000
Kpu
1000

CLint,
i t uptake
t k  CLiint,
t passive
i
CLint,passive
100
10
1
 Kpu (CLuptake/Pdiff) 250-fold range (erythromycin and atorvastatin).
Covariate analysis
y
2.5
0.5
2
0
1.5
log fuce lll
log Pdifff
-0.5
1
-1
-1.5
0.5
-2
0
-2.5
-2
0
2
Log D7.4
4
6
-2
0
2
4
Log D7.4
Useful for cross-species and cross-systems extrapolation
But no statistical relationship between:
Log D and any active uptake parameters
6
Interplay examples - Kpu = 1 :
actively transported drugs
Ki Hepa
atocytes (µM)
100
((n= 7 inhibitors, n=21 p
pathways)
y )
Inhibitor
Cell-to-Media
Ratio
(Kp)
MCZ
FXT
KCZ
FVX
QUI
OMP
FCZ
6000
2010
1200
577
143
16
4.2
10
1
Saquinavir
Enoxacin
0.1
0.01
0.01
0.1
1
10
Ki Microsomes (µM)
 Miconazole
Fluconazole
Ketoconazole
Fluoxetine
Fl
 Quinine
Fluvoxamine
Fl
 Omeprazole
Saquinavir
Nelfinavir
Ritonavir
100
 Good agreement between Ki
values in both systems (except
the higher affinity inhibitors)
 Kpu
K approximately
i t l 1
Brown et al, DMD 2007
Example
p 1:Enoxacin inhibition of
theophylline oxidation
S b t ti l DDI reported
Substantial
t d in
i humans
h
andd rats
t
Hepatocytes
Microsomes
% of control activity
THEO KM
80
THEO 2KM
60
40
20
0
No significant inhibition
Ki >1000 µM
1
10
100
[Enoxacin] (µM)
100
% of control activity
THEO 0.5KM
100
80
60
40
20
Competitive
p
inhibition
Ki = 120  65µM (n=3)
0
1000
1
10
100
1000
[Enoxacin] (µM)
 Significantly more potent (>20) inhibition in cells vs. microsomes
 Active
i uptake
k off enoxacin
i
 Similar scenario observed with erythromycin
Brown et al, 2010
Example 2: Inhibition of CYP3A by
HIV protease inhibitors –
nelfinavir and saquinavir
 Well documented examples of actively
transported drugs with substantial DDIs
 Hepatocyte-microsomal difference in Ki to be
expected
 Similar scenario to enoxacin and erythromycin?
Saquinavir metabolism and uptake
in rat
50
v (nmol/min/g liverr)
Permeability 50ml/min/g liver
40
30
Clint >350 ml/min/g liver
Metabolism
microsomes
H t t
Hepatocytes
Clint 50 ml/min/g liver
Uptake
p
hepatocytes
20
10
0
0
1
22
33
[SQV]
[SQV] M
M
Uptake is rate limiting and defines CL
44
Hepatic uptake and
microsomal:hepatocyte Km & Ki ratios
q
and nelfinavir
for saquinavir
Microsomal:hepatocellular ratio
Drug
(Kptotal)
Km
Ki
Kpu
Saquinavir (306)
0.16
0.34
6.8
Nelfinavir (3350)
0.03
0.04
5.7
Opposite effect to enoxacin & erythromycin cases
Parker et al, DMD 2008
Framework for interplay of metabolism &
t
transporters
t
on Kp
K u
(Intermediate permeability 0.1 ml/min/M cells)
Enoxacin,
erythromycin
Saquinavir,
q
,
nelfinavir
• No efflux or tissue binding
Example 3: Repaglinide-Gemfibrozil DDI
Inhibition of both hepatic uptake and metabolism
Pl
Plasma
Li
Liver
Repaglinide
Passive
X
OATP1B1
UGT
CYP2C8
X
CYP3A4
X - inhibition by GFZ and GFZ-glucuronide
DDI at both transporter and P450 level - sequential effect
Metabolites
Repaglinide uptake in human hepatocytes
4000
Contribution to ttotal uptake (%))
C
Uptak
ke Rate (pmoles
s/mg protein/miin)
100
2000
80
60
40
20
0
0
0
20
40
60
80
Repaglinide Concentration (µM)
-- Total uptake
-- Passive component (4°C)
--- Active component (simulated data)
Km = 14.1 µM
100
0.1
1
10
100
Repaglinide concentration (µM)
1000
CLuptake 5-fold greater
than the passive component
at therapeutic concentrations
IC50 44.3
3 and 77.4
4 µM for GFZ and GFZ-glucuronide
GFZ-glucuronide, respectively
Comparison of contribution (based on Clints)
off pathways
th
across in
i vitro
it systems
t
60
Hepatocytes
50
40
30
20
10
0
B
70
60
C
S9
% Contrib
bution
70
% Contribu
ution
% Contrib
bution
A
50
40
30
20
10
M2
M4
Gluc
HLM
60
50
40
30
20
10
0
0
M1
70
M1
M2
M4
Gluc
M1
M2
M4
Gluc
● CYP3A4 ● CYP2C8 ● UGT
• M2 contributed similar % in S9 and hepatocytes
y
(66%
(
in vivo)), needs aldehyde
y
dehydrogenase to drive pathway.
• 5% contribution of M2 observed in HLM  increased importance of M1 and M4
• Similar contribution of CYP3A4 and CYP2C8 in S9 and hepatocytes
Säll et al, DMD 2012
Inhibition of CYP2C8 by
Gemfibrozil in vitro
 Using repaglinide depletion
and rosiglitazone parahydroxylation
80
60
40
25
IC50 (µM)
 Competitive inhibition
(red bars) Microsomal/hepatocyte ratios
7 and 3
20
15
10
 Time dependant inhibition, with
pre-incubation (blue bars) Mi
Microsomal/hepatocyte
l/h
t
t ratio
ti
24 and 44 [greater
accumulation of glucuronide?]
5
0
Microsomes
Hepatocytes
Prediction of Repaglinide-Gemfibrozil DDI
Repagliniide AUC in the ppresence of gemffibrozil
and its gluccuronide
40
35
30
25
20
15
10
5
0
Obs
A
Predicted from
interplay model
Summary
y
• Use of rat hepatocytes provides a comprehensive
package of clearance mechanisms for PBPK
modelling to delineate intracellular events.
• Kp parameters particularly useful. Allows
resolution of cellular binding and active transport
processes.
• Unbound intracellular drug concentration needed
– consequences
q
of transporters.
p
• Certain parameters are translatable to humans
(Pdiff, fucell)
Acknowledgements
Aleksandra Galetin, David Hallifax
Jane Alder, Hayley Brown,
Carina Cantrill,, Carolina Säll,,
Alison Wilby, Yoshiyuki Yabe
Member companies of Centre for Applied
Ph
Pharmacokinetic
ki ti Research
R
h (CAPkR) Consortium
C
ti
http://www.capkr.manchester.ac.uk
Repaglinide metabolism in human
hepatocytes, S9
S and microsomes
M1
M2
UGT1A1/1A3
CYP2C8/3A4
Repaglinide - glucuronide
Repaglinide
M4
M5
M0-OH
• Previous in vitro study identified M1 and M4 as major metabolites
• M2 reported as major in vivo metabolite (66% of dose excreted in faeces and urine)
• No in vitro data available to confirm predominant role of CYP2C8 in repaglinide metabolism