Báo cáo khoa học: Modulation of P-glycoprotein-mediated multidrug resistance by acceleration of passive drug permeation across the plasma membrane potx

Modulation of P-glycoprotein-mediated multidrug
resistance by acceleration of passive drug permeation
across the plasma membrane
Ronit Regev*, Hagar Katzir*, Daniella Yeheskely-Hayon and Gera D. Eytan
Department of Biology, the Technion ) Israel Institute of Technology, Haifa, Israel
Classic multidrug resistance (MDR) is attributed to
the elevated expression of the ATP-dependent drug
efflux pumps ABCB1 [also known as P-glycoprotein
(Pgp)], ABCC1 (also known as multidrug resistance-
associated protein) and ABCG2 (also known as breast
cancer resistance protein and mitoxantrone resistance
protein), all of which belong to the superfamily of
ATP-binding cassette (ABC) transporters [1]. Efflux
mediated by ABC drug transporters leads to decreased
cellular accumulation of anticancer drugs, which is a
main cause of the limited success of the currently
applied chemotherapy regimens. Pgp, a product of the
ABCB1 (previously known as MDR1) gene, is the most
extensively studied ABC drug transporter. Pgp trans-
ports chemically dissimilar drugs that act on diverse
targets [2]. The intracellular drug concentration in
drug-resistant cells is the outcome of competition
between the active export of drugs by the efflux pumps
Keywords
anesthetics; modulators; multidrug
resistance; P-glycoprotein; propofol
Correspondence
G. D. Eytan, Department of Biology, The
Technion ) Israel Institute of Technology,
Haifa, Israel
Fax: +972 4 8225153

Tel: +972 4 8293406
E-mail:
Website:
*These authors contributed equally to this
work
(Received 22 July 2007, revised 9 Septem-
ber 2007, accepted 11 October 2007)
doi:10.1111/j.1742-4658.2007.06140.x
The drug concentration inside multidrug-resistant cells is the outcome of
competition between the active export of drugs by drug efflux pumps, such
as P-glycoprotein (Pgp), and the passive permeation of drugs across the
plasma membrane. Thus, reversal of multidrug resistance (MDR) can occur
either by inhibition of the efflux pumps or by acceleration of the drug per-
meation. Among the hundreds of established modulators of Pgp-mediated
MDR, there are numerous surface-active agents potentially capable of
accelerating drug transbilayer movement. The aim of the present study was
to determine whether these agents modulate MDR by interfering with the
active efflux of drugs or by allowing for accelerated passive permeation
across the plasma membrane. Whereas Pluronic P85, Tween-20, Triton X-
100 and Cremophor EL modulated MDR by inhibition of Pgp-mediated
efflux, with no appreciable effect on transbilayer movement of drugs, the
anesthetics chloroform, benzyl alcohol, diethyl ether and propofol modu-
lated MDR by accelerating transbilayer movement of drugs, with no con-
comitant inhibition of Pgp-mediated efflux. At higher concentrations than
those required for modulation, the anesthetics accelerated the passive per-
meation to such an extent that it was not possible to estimate Pgp activity.
The capacity of the surface-active agents to accelerate passive drug trans-
bilayer movement was not correlated with their fluidizing characteristics,
measured as fluorescence anisotropy of 1-(4-trimethylammonium)-6-phenyl-
1,3,5-hexatriene. This compound is located among the headgroups of the

phospholipids and does not reflect the fluidity in the lipid core of the mem-
branes where the limiting step of drug permeation, namely drug flip-flop,
occurs.
Abbreviations
ABC, ATP-binding cassette; CCCP, carbonyl cyanide m-chlorophenylhydrazone; MDR, multidrug resistance; Pgp, P-glycoprotein;
TMA-DPH, 1-(4-trimethylammonium)-6-phenyl-1,3,5-hexatriene; TMRM, tetramethylrhodamine methyl ester.
6204 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS
and the passive entry of drugs by permeation across
the plasma membrane [3].
One of the characteristics of the MDR phenotype
initially described by Skovsgaard [4] and later by
Tsuruo et al. [5] is that MDR can be reversed by
cotreatment of resistant cells with nontoxic concentra-
tions of hydrophobic compounds known as chemo-
sensitizers, modulators, or resistance modifiers. These
include calcium channel blockers (such as verapamil),
calmodulin antagonists and antiarrhythmia agents,
antihistamines, lysosomotropic amines, immunosup-
pressants, steroid hormones and modified steroids,
nonionic detergents, anesthetics, and various amphi-
pathic drugs.
MDR can be reversed by inhibition of the active
export of drugs either by binding of inhibitors on sites
located on the efflux pumps or by indirect inhibition
of Pgp either through modulation of its lipid environ-
ment or as a result of depletion of cellular ATP. Alter-
natively, MDR can be reversed by acceleration of the
passive permeation of the drugs through modulation
of the lipid environment in the plasma membrane.
Among the hundreds of established modulators of

Pgp-mediated MDR, there are numerous surface-active
agents capable of affecting the lipid environment of
membranes. They include nonionic detergents, such as
Tween-20 and Triton X-100 [6,7], excipients serving as
nontoxic diluents of drugs, such as Cremophor EL,
poly(ethyleneglycol) 300 and Pluronic block copoly-
mers [6,8,9], and anesthetics, such as benzyl alcohol,
chloroform and diethyl ether [10,11]. Sinicrope et al.
have shown that chloroform and benzyl alcohol inhibit
uptake of drugs into Pgp-containing vesicles [11]. Plu-
ronic P85 [12], chloroform, benzyl alcohol, Tween-20,
Nonidet P-40 and Triton X-100 [10] inhibited the
ATPase activity of Pgp located in either microsomes
or reconstituted vesicles. On the other hand, surface-
active modulators, such as Pluronic block copolymers
[13,14], benzyl alcohol, chloroform and various non-
ionic detergents [10,15] accelerate passive movement of
doxorubicin across artificial membranes. As the inhibi-
tion of Pgp ATPase activity exerted by the various
agents was more pronounced than their effect on the
flip-flop of drugs, it has been assumed that anesthetics
and nonionic detergents modulate Pgp-mediated MDR
by inhibition of Pgp [10].
The aim of the present study was to elucidate the
mechanism by which the various surface-active modu-
lators reverse MDR in Pgp-overexpressing cells:
whether they inhibit the Pgp-mediated active export,
or allow for accelerated passive uptake of drugs. The
effect of these agents on Pgp-mediated activity was
estimated as the active efflux of daunorubicin upon

replenishment of cellular ATP in drug-resistant cells.
Under these conditions, the initial efflux is due to
active transport of the anthracycline without the con-
tribution of passive transbilayer drug movement.
Whereas the nonionic detergents, excipients and Plu-
ronic block copolymers inhibited Pgp-mediated drug
export, the anesthetics modulated MDR by accelera-
tion of passive drug uptake, with no concomitant inhi-
bition of Pgp-mediated drug efflux.
Results
Effect of modulators on cellular uptake and efflux
of daunorubicin
Certain modulators of Pgp-mediated MDR, such as
detergents, excipients, and anesthetics, can affect
membrane characteristics such as fluidity. In order to
determine whether these modulators enhance drug
uptake into drug-resistant cells by inhibition of Pgp or
by allowing for accelerated passive movement of the
drugs across the plasma membrane, the active efflux
of daunorubicin upon replenishment of cellular ATP
to drug-resistant cells was measured using spectrofluo-
rometry. Measurement of the intracellular presence of
daunorubicin inside the cells was based on the obser-
vation that its fluorescence is quenched upon intercala-
tion between the DNA base pairs in the nucleus. The
affinity of binding of daunorubicin to DNA is high,
and thus cellular uptake of daunorubicin can be mea-
sured as the quenching of its fluorescence by the
nuclear DNA [16]. Daunorubicin enters rapidly into
drug-sensitive cells, presumably by passive transport

across the plasma membrane (Fig. 1, trace a). As
expected, depletion of cellular ATP did not affect dau-
norubicin uptake into these cells (Fig. 1, trace b). On
the other hand, daunorubicin uptake into drug-resis-
tant cells was very limited. Apparently, as the intra-
cellular concentration of daunorubicin rises, Pgp
prevents further net drug uptake. Thus, after an initial
period of rapid daunorubicin uptake, further uptake
was largely prevented (Fig. 1, trace c). Depletion of
cellular ATP by glucose deprivation combined with
poisoning of mitochondrial respiration by sodium
azide resulted in an intracellular ATP concentration
equal to 0.2 mm. Under these conditions, cellular Pgp
with a K
m
of 0.8 mm [17] is expected to operate at
20% of its capacity. This allows for increased uptake
of daunorubicin, but not to the levels observed in
drug-sensitive cells. Upon subsequent addition of glu-
cose to these cells, cellular ATP levels were replenished
and Pgp actively removed the excess daunorubicin
from the cells (Fig. 1, trace d).
R. Regev et al. MDR modulation by acceleration of drug permeation
FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS 6205
Inhibitors of Pgp known to interact directly with the
transporter, such as verapamil and cyclosporin A,
interfered with the Pgp-mediated efflux of daunorubi-
cin. In the presence of these modulators, the Pgp-medi-
ated efflux observed upon replenishment of glucose
was largely eliminated, and daunorubicin uptake into

these cells was similar to that observed in drug-sensi-
tive cells (Fig. 1, trace e, and Fig. 2). There was a clear
correlation between the degree of inhibition of Pgp-
mediated efflux and the increase in daunorubicin
uptake into the drug-resistant cells (Fig. 3). As
expected, these inhibitors had no effect on the uptake
of daunorubicin into drug-sensitive cells. Using a simi-
lar methodology; we studied MDR modulation by
eight membrane-active agents. The detergents Tween-
80 and Triton X-100, the excipient Cremophor EL and
the block copolymer Pluronic P85 modulated Pgp-
mediated resistance by a mechanism similar to that
observed with the Pgp inhibitors verapamil and cyclo-
sporin A. They had no effect on the passive uptake of
daunorubicin into drug-sensitive cells. They enhanced
daunorubicin uptake into drug-resistant cells as a
result, and in correlation with inhibition of the active
export of daunorubicin by Pgp (Fig. 4).
On the other hand, the general anesthetics chloro-
form, propofol, and diethyl ether, and the local
anesthetic benzyl alcohol, appeared to modulate Pgp-
mediated MDR by a different mechanism. They accel-
erated the passive uptake of the drug, rather than
interfering with the Pgp-mediated activity. These drugs
enhanced daunorubicin uptake into drug-sensitive cells
as well as into drug-resistant cells. The enhanced
Daunorubicin
Glucose
Glucose
Cells

a
b
b
c
d
f
e
Daunorubicin Fluorescence
g
30 minutes
Fig. 1. Daunorubicin uptake and efflux into and out of drug-sensi-
tive and drug-resistant K562 cells. Drug-sensitive cells (traces a and
b) or drug-resistant cells (traces c–g) were incubated for 30 min
either in a medium containing 5 m
M glucose (traces a and c) or in
ATP depletion medium containing 10 m
M sodium azide and no glu-
cose (traces b and d–g). Subsequently, 5 l
M daunorubicin (thin
arrows) and 5 l
M cyclosporin A (trace e), 0.2 mM propofol (trace f)
or propofol and cyclosporin A (trace g) were added. At the time
points marked by the thick arrows, 5 m
M glucose were added.
Daunorubicin fluorescence was monitored continuously.
Sensitive
Daunorubicin
Resistant Cells
Daunorubicin
Sensitive Cells

Glucose
0
0
Cells
5
5
10
10
Cells
20
20
0 60306030
Time [minutes]
Daunorubicin Fluorescence
Fig. 2. Verapamil effect on uptake and influx
of daunorubicin into and out of drug-resis-
tant and drug-sensitive cells. Drug-resistant
and drug-sensitive K562 cells were incu-
bated for 30 min under ATP-depleting condi-
tions in the presence of 10 m
M sodium
azide and the absence of glucose. Subse-
quently, various concentrations of verapamil
(marked next to the traces in terms of
micromolar concentrations) were added
together with 5 l
M daunorubicin. Daunorubi-
cin was taken up by the cells, and subse-
quently, at the time points marked by the
empty arrows, 5 m

M glucose was added,
allowing Pgp to expel the drug from the
drug-resistant cells. Daunorubicin fluores-
cence was monitored continuously.
MDR modulation by acceleration of drug permeation R. Regev et al.
6206 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS
uptake of daunorubicin in their presence was not cor-
related with inhibition of daunorubicin export by Pgp.
Moreover, at low concentrations, these agents even
accelerated the Pgp-mediated export of daunorubicin
(Fig. 1, trace f, and Fig. 5). As expected, the active
export from the drug-resistant cells was largely elimi-
nated by the Pgp inhibitor cyclosporin A (Fig. 1, tra-
ces e and g). When applied at higher concentrations,
the anesthetics further enhanced the rate of passive
drug permeation and interfered with the active export
of the drug by Pgp.
Effect of modulators on the efflux of the
rhodamine analog tetramethylrhodamine
methyl ester (TMRM)
Daunorubicin uptake, measured as described above,
reflects cellular uptake of the drug, its transfer into the
nucleus, and binding to the DNA. To assess the effect
of modulators specifically on drug transport across the
cell plasma membrane, we characterized the effect of
the modulators on the cellular pharmacokinetics of the
rhodamine analog TMRM. TMRM taken up by cells
accumulates in the mitochondria in response to the
mitochondrial electrochemical potential [18]. This
Fig. 4. Quantitative effects of Pluronic P85,

Tween-20, Chremophor EL and Triton X-100
on uptake and efflux of daunorubicin into
and out of drug-resistant and drug-sensitive
cells. The effects of the above-mentioned
agents on daunorubicin uptake and efflux
into and out of drug-resistant and drug-sen-
sitive cells was quantitated as described in
Fig. 3.
2.0
0.2
Daunorubicin Uptake
[nmoles]
1.0
0.1
Daunorubcin Efflux
[nmoles min
–1
]
02010
0.0 0.0
Vera
p
amil [µM]
Fig. 3. Quantitative effect of verapamil on uptake and efflux of dau-
norubicin into and out of drug-resistant and drug-sensitive cells.
The effects of verapamil on daunorubicin transport in drug-resistant
and drug-sensitive cells was measured as described in Fig. 2. As
no curve could be fitted to the kinetics of daunorubicin uptake into
the resistant cells, the uptake rate was measured as the amount of
drug taken up within 15 min after addition of daunorubicin either to

drug-resistant cells (squares) or to drug-sensitive cells (triangles).
The amount of daunorubicin pumped out of the resistant cells by
Pgp (circles) was assessed by a linear regression to the efflux
kinetics exhibited during the 3 min period subsequent to the first
minute after the addition of glucose to the cells.
R. Regev et al. MDR modulation by acceleration of drug permeation
FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS 6207
accumulation can be eliminated by dissipation of this
potential with the uncoupler carbonyl cyanide m-chlo-
rophenylhydrazone (CCCP), thus allowing for the
measurement of transport into and out of the cell cyto-
plasm without involvement of the mitochondrial sink.
Under these conditions, the transport rate of the dye is
limited by the plasma membrane.
The observed effects of the various modulators on
TMRM efflux were in accord with the results obtained
with daunorubicin. The Pgp inhibitors cyclosporin A
and verapamil, as well as Tween-80, Triton X-100,
Cremophor EL, and Pluronic P85, inhibited efflux of
TMRM preloaded into drug-resistant cells and had lit-
tle effect on efflux of TMRM preloaded into drug-sen-
sitive cells (Fig. 6 and data not shown). In contrast,
the anesthetics (chloroform, benzyl alcohol, diethyl
ether, and propofol) accelerated the efflux of TMRM
preloaded into drug-resistant and drug-sensitive cells
(Fig. 6 and data not shown). The effect was much
more pronounced in drug-sensitive cells, but was also
significant in drug-resistant cells.
In the absence of the uncoupler, the dye TMRM
accumulated in the mitochondria of K562 cells. The

amount of TMRM taken up by the cells was high, and
the time course of the efflux was longer. Thus, it was
possible to measure uptake and efflux by rapid separa-
tion of the cells from the incubation medium. All the
Pgp modulators mentioned above enhanced TMRM
uptake into resistant cells. The Pgp inhibitors cyclo-
sporin A and verapamil, and Cremophor EL, Tween-
20, and Pluronic 85, inhibited Pgp-mediated efflux of
TMRM (Fig. 7). On the other hand, the anesthetics
benzyl alcohol, chloroform and propofol accelerated
the efflux of TMRM (Fig. 7).
Lack of correlation between acceleration of drug
permeation and membrane fluidity as measured
with 1-(4-trimethylammonium)-6-phenyl-1,3,5-
hexatriene (TMA-DPH)
It has been suggested that the characteristic of certain
anesthetics and surface agents relevant to the modula-
tion of Pgp-mediated MDR is their capacity to alter
the fluidity of the cell plasma membranes [6,10,11,19].
Therefore, the effect of the above-mentioned Pgp
modulators on membrane fluidity was studied by
measuring the steady-state anisotropy fluorescence of
TMA-DPH. TMA-DPH is a short-chain lipid analog
with a hydrophilic head consisting of a constitutively
positively charged quaternary amine and a hydropho-
bic tail. This dye is localized among the headgroups of
the phospholipids located in the cell plasma membrane
Fig. 5. Quantitative effects of propofol,
chloroform, diethyl ether and benzyl alco-
hol on uptake and efflux of daunorubicin

into and out of drug-resistant and drug-
sensitive cells. The effect of the above-
mentioned agents on daunorubicin uptake
and efflux into and out of drug-resistant
and drug-sensitive cells was quantitated as
described in Fig. 3.
MDR modulation by acceleration of drug permeation R. Regev et al.
6208 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS
[20–22]. The modulators had a variable effect on the
fluorescence anisotropy of TMA-DPH, and no clear
correlation could be observed between this effect and
the efficacy of the agents in the modulation of Pgp-
mediated TMRM efflux (Fig. 8). Thus, whereas the
anesthetics chloroform, benzyl alcohol and diethyl
ether decreased the fluorescence anisotropy of TMA-
DPH, propofol had little effect on the fluorescence
anisotropy. Yet, all these anesthetics reversed Pgp-
mediated MDR by acceleration of the passive
transmembrane movement. The effect of the other
surface-active agents on membrane fluidity, measured
as TMA-DPH fluorescence anisotropy, varied between
no effect and marked effect (Fig. 8). However, none of
the other agents accelerated the transbilayer perme-
ation of the drugs and dyes.
Discussion
Analysis of the reversal of Pgp-mediated MDR by var-
ious membrane-active agents revealed two apparently
conflicting mechanisms of MDR modulation. The
detergents, excipients and Pluronic block copolymers
reversed the resistance by inhibition of Pgp-mediated

active transport. On the other hand, low concentra-
tions of the anesthetics benzyl alcohol, propofol, chlo-
roform and diethyl ether reversed the resistance by
acceleration of the passive transport of the drug or dye
across the plasma membrane. At these concentrations,
the anesthetics did not inhibit the Pgp-mediated efflux,
but even accelerated it. The acceleration of drug per-
meation across the plasma membrane was not due to
permeabilization of the cells, as the cells were not
stained with membrane-impermeable stains, such as
propidium iodide. More significantly, the cells retained
the capacity of Pgp to pump drugs out of the cells.
Interestingly, the results obtained with living cells dif-
fer from the data obtained when the effects of the vari-
ous agents were analyzed using liposomes and plasma
membrane preparations [6,10,11]. In cell-free systems,
the anesthetics had a much more pronounced effect on
Pgp ATPase activity in comparison to their effect on
the drug flip-flop across the membrane. In contrast, in
K562 cells, low concentrations of anesthetics affected
exclusively the permeation rate of the drugs. On the
other hand, the nonionic detergents accelerated flip-
flop in liposomes, but had little effect on permeation
in the cells. Only the simultaneous analysis of Pgp-
mediated efflux and permeation in the cells led to the
conclusion that the anesthetics modulate MDR by
acceleration of the permeation, whereas the other
agents modulate MDR by inhibition of the active
efflux.
Net activity of Pgp without interference of the pas-

sive permeation of the drugs was observed as daunoru-
bicin efflux from drug-resistant cells. In these
experiments, Pgp-mediated net transport was observed
after replenishment of cellular ATP levels by glucose
A
B
D
C
Fig. 6. Effects of propofol and Pluronic P85
on TMRM efflux from cells. Drug-resistant
(A, C) and drug-sensitive (B, D) K562 cells
were loaded with TMRM in the presence of
CCCP (1 l
M) and in the presence or
absence of verapamil (10 l
M), respectively.
The apparent intracellular TMRM concentra-
tions in drug-sensitive and drug-resistant
cells were about 30 and 12 times the con-
centration in the medium, respectively.
TMRM was removed by pelleting the cells
and suspending them in CCCP-containing
fresh medium in the absence (empty circles)
or presence of the anesthetic propofol at
0.1 m
M (empty squares), 0.2 mM (full
circles), or 0.5 m
M (full squares), or the
surface-active agent Pluronic P85 at
0.0001% (empty squares), 0.001% (full

circles), 0.01% (triangles), or 0.1% (full
squares). The amount of TMRM associated
with the cells was determined by flow
cytometry as described in Experimental
procedures.
R. Regev et al. MDR modulation by acceleration of drug permeation
FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS 6209
supplement. Before the addition of glucose, the cellular
content of daunorubicin was in quasi-equilibrium
resulting from equal rates of the passive inward trans-
port of the drug and the residual active drug efflux cat-
alyzed by Pgp. Upon replenishment of ATP, the initial
outward drug transport, away from this equilibrium, is
due only to the activity of Pgp. Thus, the acceleration
of the efflux observed in the presence of the anesthetics
is mediated by Pgp activity and is not due to faster
drug permeation. A plausible cause for this accelera-
tion is faster passive movement of the drugs to the
active site of the Pgp. It has been suggested that Pgp-
mediated dye or drug export involves incorporation of
the drugs into the inner monolayer of the plasma
membrane, its lateral movement toward the active site
of Pgp, and active extrusion by the latter directly from
the lipid phase of the inner leaflet of the plasma mem-
brane [23]. This model is supported by experimental
data, including observation of suitable side entrances
from the lipid matrix into the protein that may permit
access of substrates to the core of the protein [24–28].
The anesthetics could facilitate the incorporation of
drugs into the inner leaflet of the plasma membrane

and ⁄ or accelerate its lateral transport to the Pgp. Tran
et al. [29] have analyzed the kinetic parameters of Pgp-
mediated transport across a confluent monolayer of
canine cells, and concluded that the association of the
drugs with Pgp was rate-limited by their lateral diffu-
sion in the plasma membrane. In contrast to the situa-
tion when daunorubicin efflux was measured, the
anesthetic-accelerated efflux of preloaded TMRM was
due mainly to acceleration of passive transport of the
dye. In this case, the efflux is due to Pgp activity and
passive transport of the dye. However, the acceleration
of the passive transport by itself could account for the
observed enhanced efflux rate.
AB
D
C
FE
Fig. 7. Effects of various agents on TMRM
efflux from K562 cells. Drug-resistant cells
(A, C, E) and drug-sensitive K562 cells (B,
D, F) were loaded with TMRM in the pres-
ence or absence of verapamil (10 l
M),
respectively. The apparent intracellular
TMRM concentrations in drug-sensitive and
drug-resistant cells were about 200 and 40
times, respectively, the concentration in the
medium. The cells were separated from the
TMRM-containing medium by centrifugation
and suspension in fresh medium in the

absence (empty circles) or presence of the
following additions: (A, B) the Pgp inhibitors
verapamil (30 l
M, squares) or cyclospor-
ine A (10 l
M, full circles); (C, D) the surface-
active agents Cremophor EL (0.05%, empty
squares), Tween-20 (0.01%, full circles), or
Pluronic P85 (0.01%, full squares); (E, F) the
anesthetics chloroform (10 m
M, empty
squares), benzyl alcohol (10 m
M, full circles),
or propofol (0.2 m
M, full squares). Cell sam-
ples were obtained after various further
incubation periods, and the amount of
TMRM associated with them was deter-
mined after centrifugation through oil cush-
ions as described in Experimental
procedures.
MDR modulation by acceleration of drug permeation R. Regev et al.
6210 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS
Passive transbilayer movement of MDR-type drugs
and dyes involves the incorporation of the agent into
the proximal membrane leaflet, transbilayer flip-flop,
and release from the opposing leaflet. The incorpora-
tion and release of these agents is very fast, and they
are practically in equilibrium between the liquid phase
and the membranes [29,30]. Thus, the anesthetic-medi-

ated acceleration of transbilayer drug movement is due
either to enhanced rate of drug flip-flop across the lipid
core of the membranes or to greater affinity of the
drugs for the plasma membrane. Anesthetics and non-
ionic detergents have been shown to accelerate flip-flop
of doxorubicin and mitoxantrone across liposome mem-
branes [10,15]. As Breuzrd et al. have observed that
modulation of mitoxantrone resistance in Pgp-overex-
pressing cells reduces the amount of drug incorporated
in the plasma membranes [31], it seems that the anes-
thetics accelerate the permeation by enhancing the drug
flip-flop rate across the lipid core of the membrane.
Benzyl alcohol, diethyl ether and other fluidizing
agents inhibit the activity of Pgp in cell-free systems
[6,10,11]. The inhibition of Pgp activity could be due
either to a direct effect on the enzyme or to an indirect
effect mediated by alterations in membrane structure.
As the inhibitory concentrations of fluidizers are simi-
lar to those required for membrane fluidization, it has
been suggested that Pgp-mediated modulation of
MDR is due to increased membrane fluidity, evident
as reduced fluorescence anisotropy of probes, such as
DPH [6,11]. However, Dudeja et al. have observed
that polyoxyethylene surfactants reverse MDR in KB-
8-5-11 drug-resistant cells by decreasing membrane flu-
idity measured as fluorescence anisotropy of a variety
of membrane probes [32]. Similarly, Woodcock et al.
[6] have ascribed modulation of MDR by Chremo-
phor EL to a reduction in membrane fluidity. Hugger
et al. evaluated the activity of Pgp in cell monolayers

and found no correlation between the inhibition of
Pgp activity and fluidity of the membranes measured
as fluorescence anisotropy of TMA-DPH [33]. In the
present study, there was no correlation between the
capacity of the various membrane-active agents to
reverse MDR and their effect on membrane fluidity as
measured with TMA-DPH. TMA-DPH is localized
among the headgroups of the phospholipids at the
outer surface of the plasma membrane [20–22]. The
membrane characteristics relevant to the effect of
the anesthetics on MDR are probably located at the
hydrophobic core of the membrane, where drug flip-
flop across the membrane occurs and the side
entrances of Pgp are located [28].
The apparent contradiction between the anesthetic-
mediated inhibition of Pgp ATPase activity in cell-free
systems and their mode of modulation in K562 cells
could be due to the combined effect of the anesthetics,
inhibition of Pgp activity and acceleration of passive
drug permeation. In living cells, the anesthetics acceler-
ate drug permeation at concentrations lower than
Fig. 8. Effect of the various anesthetics on membrane lipid fluidity measured with TMA-DPH. Anisotropy (r) values were measured by
steady-state fluorescence polarization at 25 °C using the fluorescent probe TMA-DPH.
R. Regev et al. MDR modulation by acceleration of drug permeation
FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS 6211
those required to inhibit Pgp ATPase activity. Like-
wise, the agents tested here, other than the anesthetics,
inhibited Pgp-mediated efflux and had no effect on the
rate of passive transbilayer movement. Yet, it has been
shown that diblock polymers accelerate movement of

drugs, including doxorubicin, across artificial mem-
branes [13]. At higher concentrations and in other
cells, these agents do accelerate the transbilayer move-
ment of drugs. Thus, diblock polymers, such as Plu-
ronic P85, inhibited Pgp-mediated transport across cell
monolayers, but had no effect on the passive transbi-
layer movement of drugs. Higher concentrations of the
copolymers accelerated the passive transport across the
cell monolayers [34,35]. A similar conclusion was
reached by Bogman et al., who measured rhoda-
mine 123 efflux from leukemia cells [36]. On the other
hand, poly(ethylene glycol) modulated resistance in
Caco-2 cell monolayers by concomitant inhibition of
Pgp-mediated active transport and acceleration of pas-
sive movement of drugs [33].
The capacity of anesthetics to accelerate passive
drug permeation without membrane permeabilization
could play a role in various functions besides modula-
tion of MDR, such as modulation of pharmacokinetics
in anesthetized patients, especially enhancement of pas-
sive transport of drugs across the blood–brain barrier.
The present study demonstrates that anesthetics modu-
late MDR by accelerating passive drug permeation.
This mechanism of MDR modulation is not necessar-
ily limited to anesthetics, and could be operative in
MDR modulation by other modulators and under
other circumstances.
Experimental procedures
Daunorubicin, TMRM, TMA-DPH, 12-AS and mineral oil
were purchased from Sigma (Rehovot, Israel). K562, a

human leukemia cell line established from a patient with a
chronic myelogeneous leukemia in blast transformation [37],
was purchased from ATCC (Rockville, MD) and main-
tained in Iscove’s medium supplemented with 100 lgÆmL
)1
penicillin G, 100 lgÆmL
)1
streptomycin, 2 mm glutamine,
and 10% fetal bovine serum (Biological Industries,
Beit-haemmek, Israel). The K562 Pgp-overexpressing sub-
line was obtained by sequential exposure of cells to increas-
ing concentrations of doxorubicin. The resistant subline was
maintained in the presence of 0.5 lm doxorubicin.
Continuous monitoring of daunorubicin cellular
transport
Uptake and efflux of the anthracycline daunorubicin was
assayed by monitoring the quenching of the drug’s fluores-
cence upon intercalation between base pairs of the nuclear
DNA, essentially as previously described [38]. The drug flu-
orescence was monitored continuously in a Cary eclipse flu-
orescence spectrophotometer (Varian Inc., Palo Alto, CA),
with the temperature kept at 37 °C. In a typical experiment,
10
6
cells were partially depleted of ATP by incubation with
stirring for 30 min in 2 mL of medium composed of NaCl
(132 mm), KCl (3.5 mm), CaCl
2
(1 mm), MgCl
2

(0.5 mm),
sodium azide (1 mm), and Hepes ⁄ Tris buffer (20 mm,
pH 7.4). Subsequently, the agent whose effect was being
tested was added, and after a further 5 min of incubation,
daunorubicin (5 lm) was added. Finally, cellular ATP was
replenished by addition of glucose (10 mm). In experiments
not involving ATP depletion, the medium contained glucose
(10 mm) instead of sodium azide. Daunorubicin fluores-
cence was monitored using an excitation wavelength of
490 nm and an emission wavelength of 595 nm. The extent
of fluorescence quenching upon binding of daunorubicin to
DNA was determined by exposure of daunorubicin to satu-
rating amounts of DNA, which resulted in quenching of
94% of the fluorescence. The quenching of daunorubicin
bound to cellular DNA was converted to amount of dauno-
rubicin by using a factor equal to the fluorescence of an
equivalent daunorubicin solution in the absence of cells
divided by the amount of daunorubicin multiplied by 0.94.
Measurement of TMRM efflux from cells
TMRM efflux was monitored in cells whose mitochondrial
membrane potential was dissipated by CCCP. Under these
conditions, the dye efflux was rapid and was monitored by
flow cytometry. K562 cells were loaded with TMRM as
follows: the cells were washed once with a medium composed
of Dulbecco’s NaCl ⁄ P
i
supplemented with CCCP (1 lm),
MgCl
2
(1 mm), CaCl

2
(1 mm), and glucose (10 mm), sus-
pended to a density of 5 · 10
5
cellsÆmL
)1
in the same med-
ium, also containing TMRM (10 lm), and incubated for
30 min at 37 °C. Cell aliquots were pelleted and kept on ice
until the assay was begun by their suspension in dye-free
medium pre-equilibrated at 15 °C. Samples of 2000 cells each
were analyzed by flow cytometry measurements in a Becton
Dickinson (Franklin Lakes, NJ) FACScan flow cytometer
equipped with an argon ion laser and a thermostated jacket
of the assay tube. The fluorescence signal was detected
through the standard FL2 channel, and was gated using the
forward and side scatterings to exclude dead cells and debris
from the analysis.
TMRM efflux from cells not exposed to the uncoupler
CCCP was measured by a quantitative assay based on the
rapid separation of the cells from the external medium.
Apart from the absence of CCCP, the experimental setup
was similar to that described above. To measure the
TMRM amount associated with the cells, 0.4 mL samples
were withdrawn and placed in an Eppendorf-style microfuge
MDR modulation by acceleration of drug permeation R. Regev et al.
6212 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS
above a 0.2 mL cushion composed of 95 parts of silicone oil
AR 200 (d
20

¼ 1.049) and five parts of mineral oil (d
20
¼
0.89). After centrifugation for 4 min at 16 060 g at room
temperature, the oil cushion was washed three times with
water by suction, and all the upper phase, including part of
the oil cushion but leaving a fraction of the oil above the
cell pellets, was removed. The cell pellets were dissolved by
addition of 0.05 mL of guanidine hydrochloride (5 m) buf-
fered with Hepes ⁄ Tris (50 mm, pH 7.4), centrifugation for
5 min, and incubation for at least 1 h at room temperature.
The dissolved samples were mixed thoroughly with 0.5 mL
of water and centrifuged for 5 min. Samples (0.4 mL) were
withdrawn from the pellets dissolved in the aqueous phase.
The TMRM fluorescence was determined using an excita-
tion wavelength of 552 nm and an emission wavelength of
580 nm. To ensure fidelity of the assay, dye-free cell samples
were mixed with known amounts of TMRM and processed
as above. The TMRM yield thus obtained matched the
amount expected. It was determined that the time required
to separate cells from the external medium was 1.5 min. All
curves were adjusted accordingly.
Membrane fluidity measurements
K562 cells were labeled with TMA-DPH (1 lm) by incuba-
tion for 5 min at 24 °C. Steady-state fluorescence anisotropy
(r) was determined using excitation and emission wave-
lengths of 358 nm and 428 nm, respectively. Fluorescence
anisotropy was calculated as described by Shinitzky &
Barenholz [39]. According to these authors, the fluorescence
anisotropy values are inversely proportional to cell mem-

brane fluidity. A high degree of fluorescence anisotropy rep-
resents a high structural order or low cell membrane fluidity.
Studies on membrane fluidity performed with the fluorescent
probe TMA-DPH yielded information about the rigidity of
the cell membrane near the lipid polar heads [40]. All the flu-
orescence measurements were repeated six times and cor-
rected for the contribution of light scattering by performing
control experiments on cells without fluorescent probes.
References
1 Gottesman MM, Fojo T & Bates SE (2002) Multidrug
resistance in cancer: role of ATP-dependent transport-
ers. Nat Rev Cancer 2, 48–58.
2 Ambudkar SV, Kimchi-Sarfaty C, Sauna ZE & Gottes-
man MM (2003) P-glycoprotein: from genomics to
mechanism. Oncogene 22, 7468–7485.
3 Eytan GD, Regev R, Oren G & Assaraf YG (1996) The
role of passive transbilayer drug movement in multidrug
resistance and its modulation. J Biol Chem 271, 12897–
12902.
4 Skovsgaard T (1980) Circumvention of resistance to
daunorubicin by N-acetyldaunorubicin in Ehrlich ascites
tumor. Cancer Res 40, 1077–1083.
5 Tsuruo T, Iida H, Tsukagoshi S & Sakurai Y (1981)
Overcoming of vincristine resistance in P388 leukemia
in vivo and in vitro through enhanced cytotoxicity of
vincristine and vinblastine by verapamil. Cancer Res 41,
1967–1972.
6 Woodcock DM, Linsenmeyer ME, Chojnowski G, Krie-
gler AB, Nink V, Webster LK & Sawyer WH (1992)
Reversal of multidrug resistance by surfactants. Br J

Cancer 66, 62–68.
7 Schuldes H, Dolderer JH, Zimmer G, Knobloch J, Bick-
eboller R, Jonas D & Woodcock BG (2001) Reversal of
multidrug resistance and increase in plasma membrane
fluidity in CHO cells with R-verapamil and bile salts.
Eur J Cancer 37, 660–667.
8 Hugger ED, Novak BL, Burton PS, Audus KL & Borc-
hardt RT (2002) A comparison of commonly used poly-
ethoxylated pharmaceutical excipients on their ability to
inhibit P-glycoprotein activity in vitro. J Pharm Sci 91,
1991–2002.
9 Kabanov AV, Batrakova EV & Alakhov VY (2002)
Pluronic block copolymers as novel polymer therapeu-
tics for drug and gene delivery. J Control Release 82,
189–212.
10 Regev R, Assaraf YG & Eytan GD (1999) Membrane
fluidization by ether, other anesthetics, and certain
agents abolishes P-glycoprotein ATPase activity and
modulates efflux from multidrug-resistant cells. Eur J
Biochem 259, 18–24.
11 Sinicrope FA, Dudeja PK, Bissonnette BM, Safa AR &
Brasitus TA (1992) Modulation of P-glycoprotein-medi-
ated drug transport by alterations in lipid fluidity of rat
liver canalicular membrane vesicles. J Biol Chem 267,
24995–25002.
12 Batrakova EV, Li S, Li Y, Alakhov VY & Kabanov
AV (2004) Effect of pluronic P85 on ATPase activity of
drug efflux transporters. Pharm Res 21, 2226–2233.
13 Erukova VY, Krylova OO, Antonenko YN & Melik-
Nubarov NS (2000) Effect of ethylene oxide and propyl-

ene oxide block copolymers on the permeability of
bilayer lipid membranes to small solutes including doxo-
rubicin. Biochim Biophys Acta 1468, 73–86.
14 Krylova OO, Melik-Nubarov NS, Badun GA, Kseno-
fontov AL, Menger FM & Yaroslavov AA (2003)
Pluronic L61 accelerates flip-flop and transbilayer doxo-
rubicin permeation. Chemistry 9, 3930–3936.
15 Regev R, Yeheskely-Hayon D, Katzir H & Eytan GD
(2005) Transport of anthracyclines and mitoxantrone
across membranes by a flip-flop mechanism. Biochem
Pharmacol 70, 161–169.
16 Mankhetkorn S, Dubru F, Hesschenbrouck J, Fiallo M
& Garnier-Suillerot A (1996) Relation among the resis-
tance factor, kinetics of uptake, and kinetics of the
P-glycoprotein-mediated efflux of doxorubicin, daunoru-
bicin, 8-(S)-fluoroidarubicin, and idarubicin in multi-
drug-resistant K562 cells. Mol Pharmacol 49, 532–539.
R. Regev et al. MDR modulation by acceleration of drug permeation
FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS 6213
17 Urbatsch IL, al-Shawi MK & Senior AE (1994) Charac-
terization of the ATPase activity of purified Chinese
hamster P-glycoprotein. Biochemistry 33, 7069–7076.
18 Eytan GD, Regev R, Oren G, Hurwitz CD & Assaraf
YG (1997) Efficiency of P-glycoprotein-mediated exclu-
sion of rhodamine dyes from multidrug-resistant cells is
determined by their passive transmembrane movement
rate. Eur J Biochem 248, 104–112.
19 Kabanov AV, Batrakova EV & Alakhov VY (2002)
Pluronic block copolymers for overcoming drug resis-
tance in cancer. Adv Drug Deliv Rev 54, 759–779.

20 Engel LW & Prendergast FG (1981) Values for and sig-
nificance of order parameters and ‘cone angles’ of flu-
orophore rotation in lipid bilayers. Biochemistry 20,
7338–7345.
21 Kuhry JG, Fonteneau P, Duportail G, Maechling C &
Laustriat G (1983) TMA-DPH: a suitable fluorescence
polarization probe for specific plasma membrane fluidity
studies in intact living cells. Cell Biophys 5, 129–140.
22 Prendergast FG, Haugland RP & Callahan PJ (1981)
1-[4-(Trimethylamino)phenyl]-6-phenylhexa-1,3,5-triene:
synthesis, fluorescence properties, and use as a fluores-
cence probe of lipid bilayers. Biochemistry 20, 7333–7338.
23 Higgins CF & Gottesman MM (1992) Is the multidrug
transporter a flippase? Trends Biochem Sci 17, 18–21.
24 Hollo Z, Homolya L, Davis CW & Sarkadi B (1994)
Calcein accumulation as a fluorometric functional assay
of the multidrug transporter. Biochim Biophys Acta
1191, 384–388.
25 Raviv Y, Pollard HB, Bruggemann EP, Pastan I & Got-
tesman MM (1990) Photosensitized labeling of a func-
tional multidrug transporter in living drug-resistant
tumor cells. J Biol Chem 265, 3975–3980.
26 Rosenberg MF, Velarde G, Ford RC, Martin C, Ber-
ridge G, Kerr ID, Callaghan R, Schmidlin A, Wooding
C, Linton KJ et al. (2001) Repacking of the transmem-
brane domains of P-glycoprotein during the transport
ATPase cycle. EMBO J 20, 5615–5625.
27 Qu Q & Sharom FJ (2002) Proximity of bound Hoechst
33342 to the ATPase catalytic sites places the drug bind-
ing site of P-glycoprotein within the cytoplasmic mem-

brane leaflet. Biochemistry 41, 4744–4752.
28 Higgins CF & Linton KJ (2001) Structural biology. The
xyz of ABC transporters. Science 293, 1782–1784.
29 Tran TT, Mittal A, Aldinger T, Polli JW, Ayrton A,
Ellens H & Bentz J (2005) The elementary mass action
rate constants of P-gp transport for a confluent mono-
layer of MDCKII-hMDR1 cells. Biophys J 88, 715–738.
30 Abreu MS, Estronca LM, Moreno MJ & Vaz WL
(2003) Binding of a fluorescent lipid amphiphile to albu-
min and its transfer to lipid bilayer membranes. Biophys
J 84, 386–399.
31 Breuzard G, Piot O, Angiboust JF, Manfait M, Candeil
L, Del Rio M & Millot JM (2005) Changes in adsorp-
tion and permeability of mitoxantrone on plasma mem-
brane of BCRP ⁄ MXR resistant cells. Biochem Biophys
Res Commun 329, 64–70.
32 Dudeja PK, Anderson KM, Harris JS, Buckingham L
& Coon JS (1995) Reversal of multidrug resistance phe-
notype by surfactants: relationship to membrane lipid
fluidity. Arch Biochem Biophys 319, 309–315.
33 Hugger ED, Audus KL & Borchardt RT (2002)
Effects of poly(ethylene glycol) on efflux transporter
activity in Caco-2 cell monolayers. J Pharm Sci 91,
1980–1990.
34 Batrakova EV, Li S, Miller DW & Kabanov AV (1999)
Pluronic P85 increases permeability of a broad spectrum
of drugs in polarized BBMEC and Caco-2 cell monolay-
ers. Pharm Res 16, 1366–1372.
35 Zastre J, Jackson J & Burt H (2004) Evidence for mod-
ulation of P-glycoprotein-mediated efflux by methoxy-

polyethylene glycol-block-Polycaprolactone amphiphilic
diblock copolymers. Pharm Res 21, 1489–1497.
36 Bogman K, Erne-Brand F, Alsenz J & Drewe J (2003)
The role of surfactants in the reversal of active trans-
port mediated by multidrug resistance proteins. J Pharm
Sci 92, 1250–1261.
37 Lozzio CB & Lozzio BB (1975) Human chronic myelo-
genous leukemia cell-line with positive Philadelphia
chromosome. Blood 45, 321–334.
38 Tarasiuk J, Frezard F, Garnier-Suillerot A & Gattegno
L (1989) Anthracycline incorporation in human lympho-
cytes. Kinetics of uptake and nuclear concentration.
Biochim Biophys Acta 1013, 109–117.
39 Shinitzky M & Barenholz Y (1978) Fluidity parameters
of lipid regions determined by fluorescence polarization.
Biochim Biophys Acta 515, 367–394.
40 Kuhry JG, Duportail G, Bronner C & Laustriat G
(1985) Plasma membrane fluidity measurements on
whole living cells by fluorescence anisotropy of tri-
methylammoniumdiphenylhexatriene. Biochim Biophys
Acta 845, 60–67.
MDR modulation by acceleration of drug permeation R. Regev et al.
6214 FEBS Journal 274 (2007) 6204–6214 ª 2007 The Authors Journal compilation ª 2007 FEBS