close

Вход

Забыли?

вход по аккаунту

?

Drug Membrane Interaction and the Importance for Drug Transport Distribution Accumulation Efficacy and Resistance.

код для вставкиСкачать
60 1
Drug Membrane Interaction
Reviews and Trends
Drug Membrane Interaction and the Importance for Drug nansport,
Distribution, Accumulation, Efficacy and Resistance+)
Joachim K. Seydel'), Eugene A. Coats, Hans Peter Cordes, and Michael Wiese
Borstel Research Institute, Med.-Pharmaceut.Chemistry, D-23845 Borstel, Germany
Received March 16,1994
In this review of our and other recent work the possible role
of membranes on drug transport, distribution, accumulation,
efficacy and resistance is discussed.
The interest in drug development, research and QSAR is
especially focussed on the interaction of ligand (drug)
molecules and the proteins of the specific receptor. It is
widely assumed that the biological activity of drugs (like
tranquilizers, anesthetics, P-blockers, antidepressants, etc.)
arises as a result of binding to active sites in membrane
bound proteins, while the lipid background is considered to
play a more passive role. In the last decade, however, evidence is increasing that we may have underestimated the
influence of drug membrane interactions on drug action.
These membranes do not consist of lipids only but possess
polarized phosphate groups and neutral, positively or negatively charged head groups and are highly structured and
I
'5
chiral. The consequen
brane interaction
transport, distributi
tance cannot always
titioning processes,
In case of chemotherapeutics the interaction of the d i g
molecules with membranes of e.g. bacteria or fungi must be
considered additionally. These membranes have a much
more complex construction than mammalian membranes. In
many cases they are asymmetric as e.g the membrane of
gram negative bacteria (Fig. 1).
This membrane is characterized by an outer hydrophilic
core, rich in polysaccharides and a hydrophobic phospholipid bilayer. Therefore, a balance in hydrophobichydrophilic properties is indicated for an effective inhibitor.
Experimental work on artificial lipid layers or model
membranes has demonstrated that the structural propenies
Gram-negative Bacteria
Ong A Proldn
I tLlpopolysocchorldc
Phopkollpld
Llpoproleh
Peplidoplycon
I
Prololn
Figure 1. Representation of E . coli cell wall
+) Partly presented at the II Medicinal Chemistry Conference, Bad Nauheim, 1993.
*) To whom all correspondence should be. addressed.
Arch. Pharm. (Weinheim) 327,601-610 (1994)
0 VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1994 0365-6233/94/1010-0601 $5.00 + .25/0
/
602
Seydel, Coats, Cordes, and Wiese
Table 1
Drug Membrane t)Interaction
Action of the Membrane on D m Moleculeai
Difluaionh u g h mornbrano may -me
rate limltlng
Membrane may prevent completely diClwion b mtiw ate
Membrane may bind drug ~ M O U E I ~ ~ ~ ~ ~ )
Sdvation d drug in memkana
k.d to 0anlorrutfOn.l chuyaa d drug adecdw
P.
n
X
V
n
X
Y
U
n
U
Figure 2. Various possibilities in rate limiting steps in the binding of drugs
to proteins embedded in membranes (taken from'))
of these membranes may strongly be affected by the presence of membrane associated molecules. Some of the possible events which can amve from drug membrane interaction are summarized in Table 1.
Examples for these events will be given during this
review. A general model has been discussed by Herbette
and coworkers. It is summarized in Fig. 2l). Since binding
sites of some receptors are embedded in the membrane,
drug molecules need to interact favourably with membrane
components in order to get excess to the receptor. In addition to the influence of the membrane on drug distribution
and rate of diffusion, the change induced in the organization of the bilayer by drug molecules can affect the conformation of membrane associated receptors thus modifying
agonist-receptor interaction. Alternatively, the conformation of the drug may be changed in the membrane, or equilibrium between conformers is changed.
How can we measure and quantify drug-membrane interaction?
Because of the physicochemical properties of phospholipids, they readily form bilayers (artificial membranes, liposomes) which can be used to study drug membrane interaction and diffusion. The importance of neutral and charged
liposomes for modelling such interactions in various tissues
has been shown on numerous examples.
Methods to measure and quantify such interactions are
summarized in Fig. 3.
NMR techniques, differential scanning calorimetry, x-ray
diffraction, Ca++-displacement, fluorescence and many
other techniques are excellent tools that allow sensitive
determinations of the interaction of drugs with artificial
membranes2).
Amphiphilic drugs in particular can interact strongly with
membranes. A large number of drug molecules possess
amphiphilic properties while producing a wide variety of
pharmacological effects. These effects may be direct results
of interaction with proteins embedded in the membrane or
the indirect result of changes induced in the organization of
the phospholipids. The latter effect can, for example, lead
to changes in phase transition temp. at which transition
from the gel (p)- to liquid crystalline (a)-phase occurs
Arch. Pharm. (Weinheim) 327,601-610 (1994)
603
Drug Membrane Interaction
Methods for Studying Drug Membrane Interaction
* N M R
Differential &anning Calorimetry
Displacement of %a++ hprn Lipid b¶onolqmm
FT-IR
Fluorescentx
x.RayDimPction
In,
ESPR
MoleculuModelling
Figure 3. Methods for studying drug-membrane interaction
32
34
36
38
40
42
44
46
NMR techniques offer even more detailed informations.
NMR spectra can be characterized by the magnetic field at
which resonance occurs, depending on the surrounding ‘of
the nucleus considered, the degree of spin-spin coupling
produced by neighbouring effects, the spin lattice relaxation
and the spin-spin relaxation rate IDz,the latter
rate
expressed as line width of the resonance signal.
During the interaction between drug and phospholipid
molecules one or several of these parameters can change in
a manner characteristic of both the drug and the receptor
molecules.
Additional information on molecular conformation can be
obtained by the Nuclear Overhauser Effect (NOE) or transfer NOE used to measure the distance between
48
Oc
Phase transition of DPPC before (0)
and after (0)
drug (ratio 0.1 : 1.0 VM)
Figure 4. Phase transition of DPPC before (black) and after (grey) drug
additon (ratio 0.1 :1.O pM)
4.3
Ca
Ca
Ca
Ca
Ca
Figure 5. Schematic representation of a 4SCa”+-displacementexperiment
(taken from4))
which is characteristic for each phospholipid. This is schematically shown in Fig. 4. The effect of such drugs on the
displacement of Ca++ from phospholipid monolayers is
shown in Fig.
5374).
Arch. Pharm. (Weinheim) 327,601-610 (1994)
4.2
4.t
4.0
Dm
Figure 6. Line width broadening of the -CH2-methylene protons of 3 3 dichloro-N-butylbenzylamine as a function of lecithin concentration (taken
from3))
These methods allow the changes in membrane organization to be localized and quantified and they also provide
information on substructures of the drug molecule which
are involved in the interaction. Proton and 13C-NMR spectra in the presence of phospholipids can lead to changes in
lD, and Inzrelaxation rates. The change in lnz
is related
to a decrease in rotational freedom of drug molecules in the
presence of a “receptor”. These changes, expressed as function of the drugflipid ratio can be used to determine KD
values to quantify the degree of interaction. This parameter
can then be used for the derivation of structure-activity
(binding) relationships (Figures 6, 7)7).
604
Seydel, Coats, Cordes, and Wiese
the MW becomes > 300. The pore size of E. coli membranes is about 600 KD. If the activity is determined against
whole cell bacteria, the variation in inhibitory activity
(MIC)can be described by the steric effect of o-substituents
and the MW or the surface of the substituents, respectively
(Fig. 8).
In contrast, the inhibitory activity of the same sulfonamides against the isolated target enzyme depends solely on
the steric effect of the substituents whereas MW is not any
longer a limiting factor9).
In case of mammalian membranes it has been shown that
the placental transfer ratio (TR) of a heterogenous set of
compounds through the placental membrane can significantly be described by the octanol/water partition coefficient, i.e. only the overall partition coefficient of the molecule in octanol mimics drug transport'O).
v1,* IHzl CH2 protons
4-
3-
2I
1
~
.
'
0
~
I
0.01
'
.
"
1
.
'
0.02
'
'
1
-
.
-
-
1
0.04
0.03
log TR = 0.39 (0.068)log P - 0.478 (0.054) log (80.53P +1)
Lecithin Imgl
Figure 7. Plot of the change in lnz[Hz]for the rnethylene protons as
shown in Fig. 6 in the presence of increasing amounts of lecithin added
(taken from3))
Drug Transport
Drug transport may occur by passive diffusion processes
through bilayers or pores. In the first case we can expect to
describe this by the octanol/water partition coefficient and
in the second case by molecular weight (MW) or surface of
the drug molecule. This has been examplified by the diffusion controlled onset of inhibition of E. coli cultures by
rifampicin derivatives. The delay in onset of inhibition was
a function of lipophilicity and concentration').
On a series of sulfonamides it can be shown that MW
becomes the rate limiting step for the inhibitory activity if
Example: Pore Diffusion
QSAR of A n t i b a c M d Activity (ant) a g b t
a)
b)
whole d & c o l i
Isolated target enzyme (SYnthasd
for a senen of substituted 6 - m a f ~ a m i d o - 1 - p h 8 n y l p ~ l ~
- 0.03 (0.11)
n = 20
-
&&I7 (0.008) MW
log UMIC I -0.189 (0.02)
n I18 r = 0.m predictive power 3 = 0.88
b)
log UI,= -0.126 (0.02)
0.487
n = 16 r I O.SI
predictiva power 3 I0.93
s = 0.09
F=54
If other than bulk effects as for example hydrogen binding
capacity is involved in the diffusion process the octanol/
water partition coefficient alone is not a sufficient parameter. This is true for the transport across the more complicated blood/brain barrier. In this case Ganellin and coworkers' ') have introduced A log P, the difference between
the partition coefficient in the system octanol/water and
cyclohexan/water as a measure of overall hydrogen binding
capacity. A satisfying relation between A log P and the rate
of brain penetration for a heterogenous series of compounds
has been found.
Drug Distribution and Accumulation
There is another important aspect of drug membrane
interaction, the distribution or accumulation of drug molecules in membranes. Fig. 9 shows a scheme of log dose/
response curves that may be obtained when one ligand
interacts with two different biological samples, one contain-
I
a)
(1)
r = 0.96
LOG DOSE / RESPONSE CURVES
+ 6.01 (2.13)
-
Figure 8. Comparison of the QSAR equations for describing cell-free and
whole cell activities of sulfonamides. Molecular weight (MW) becomes
rate limiting for whole cell activities. Steric effects (MRo) are important
for drug receptor interaction (taken from'))
x
LOG Co (LIGANDI
Figure 9. Schematic log dose/response curves for one agonist reacting
with preparations of two different receptor subtypes, I and I1 (taken
from 12))
Arch. Pharm. (Weinheim) 327,601-610 (1994)
0 5
Drug Membrane Interaction
ing the receptor subtype I, the other type 11. From such
curves apparent dissociation constants of the ligand-receptor complex can be obtained.
The usual interpretation is that the stronger affinity at side
I1 is a consequence of the ligand structure meeting the
receptor requirements of site I1 better than of I.
An alternative interpretation, however, is that the ligand is
present near site I1 in a greater concentration and/or in a
more suitable conformation and orientation than near site
I"). This could be the consequence of a ligand-membrane
interaction near the membrane bound receptor. In that case
the membrane requirements of site I and I1 are different!
This can be shown by the partition behaviour of dihydropyridine type compounds into biological membranes. The
partitioning into sarcoplasmic reticulum for some derivatives has been determined by Herbette and coworker^'^) and
compared with partition coefficients obtained for the
system octanol/water (Table 2)
Table. 2: Partition coefficients of dihydropyridine-type drugs into
biological membranes and into octanol (taken from 13))
*
Table 3 (data taken from4')
Erythrocyte (EM)and tissue (T/M) medium ratios of the drugs in human
erythrocytes and rat aortas and left atria at pH 7.2 and at a medium concentration of l o 6 mom
(The valus represent mean f S.D.(n = 6))
E/M
Drugs
Flunarizine
R 56865
Nitrendipine
Verapamil
Diltiazem
Lidocaine
196
37
18
6
3
1.5
*
Arch. Pharm. (Weinheim)327,601-610 (1994)
3
682
230
71
23
I1
4.5
f 2
f 1
f 1
f 0.3
f 0.5
f 52
f 47
f 2
f 2
f 0.8
f 1.1
784
402
96
57
28
5.0
f 155
f 37
f 4
k 6
f 2
f 0.5
cholesterol is also changing membrane properties and by
this drug distribution into membranes, this effect has also to
be considered for disease states, like hypercholesterolemia.
Scheufler and coworker^'^) have published the partition
coefficients using biological membranes for another series
of compounds (Table 3). The partition coefficients of these
drugs into biological membranes correlate - as shown in
previous tables - only on a statistically not highly significant level with log Poctmol.The prediction power is very
poor (f= 0.27). In contrast a highly significant correlation
is obtained if the change in spin-spin relaxation rate
(A 1D2)of protons of the ligands in the presence of phospholipids is used as a measure of drug-membrane interaction (atria/medium AT/M).
The values shown were obtained using sarcoplasmic reticulum,
but similar values were obtained with cardiac sarcolemmal lipid
extracts. The results indicate a primary interaction of the drug
with the membrane bilayer component of these biological
membranes.
Two things are becoming immediately obvious:
1. The partitioning into biological membranes is much
stronger as compared to the octanoVwater system
2. The ranking is different
Especially the first point is of great importance. It means
that the concentration within the membrane is considerably
higher as compared to the surrounding water phase. This
means that KD values using concentrations in the water
phase are certainly not correct.
It is obvious that the partition coefficients do not only
depend on the structure of the drug molecules but also on
the structure and composition of the biological membranes.
There are data available from the literature, which strongly
~)
support these arguments. Valdivia and C o r ~ n a d o 'have
used a calcium channel reconstituted into planar bilayers of
different phospholipid constitution. They were able to show
that the affinity of a positively charged dihydropyridine is
increasing by a factor of 10 when the ratio of charged to
uncharged phospholipids was increased by a factor of 2. As
T/M (left atria)
T/M (aortas)
AMIODARONE
* *
F
-
i
T
J
O
f
V
~
J
P
'
CH3
MB865
CHSO
"""&f%
VEWAMlL
ocs3
*
*
OCHS
Figure 10. Changes in ID, for various spin systems of catamphiphilic
drugs indicated by * in the presence Of lecithin
606
Seydel, Coats. Cordes, and Wiese
log AT/M = 0.861 (0.33) log In2 + 1.428 (0.34)
(2)
n = 5 ? = 0.96 s = 0.21 r2c,o,,vd.= 0.89 F = 68.8
The predictive power is on the 90% level using the leave
one out procedure (9cross Val.)
Additional information on the substructure involved in the
interaction can be obtained by NMR measurements. In case
of verapamil a clear cut difference in involvement of substructures can be derived from the NMR measurements
(Fig. 10). Involved spin systems are indicated by an asterisk. In contrast amiodarone shows the involvement of all
substructures of the molecule in the drug lipid interaction.
This finding is supported by results from x-ray diffraction
measurements showing amiodarone deeply bumed into the
lipid hydrocarbon hai in'^^'^). Simulation of the binding
mode of the energy minimized structures of these two drugs
into a complex of 4 lipid PPC molecules are in agreement
with the x-ray and NMR studies. A high binding energy is
found for amiodarone burried between the hydrocarbon
chain whereas for verapamil the calculated binding energy
is significantly larger in the extended conformation where
that part of the molecule for which no change in l/T* has
been observed, is not interacting with the phosphor head
group but reaches into the water phase.
An example where even the preferred conformation of the
drug molecule is changed in lipid environment are N-alkyl
substituted benzylamines. These compounds produce negative inotropic effects (beating rate and contractile force,
guinea-pig atria) which increases strongly at chain length
larger than 4. We were able to show that this non-linear
increase in activity with chain length was due to a change in
preferred c~nformation~.~).
Up to a chain length of n-butyl
the extended molecular structure exists in the absence and
presence of phospholipids. With longer chain length, however, a folded conformation is preferred in the presence of
Thus the active conformation is created in lipid
environment.
Next some examples are presented where drug membrane
interaction seems to be directly related to drug efficacy.
Neuroleptics act on the sodium/potassium channels, which
are embedded on phospholipid containing membranes. In a
cooperative study with Asta Medica with the aim to reduce
animal experiments, the interaction of flupirtin derivatives
with phospholipid vesicles using NMR technique has been
analyzed. The interaction was quantified as discussed
before by the change in 1/T2for the methylene protons as a
function of increasing phospholipid concentration. Linear
dependence was observed for a large range of lipid concentrations. The slope of such plots was used to characterize
the degree of interaction. A highly significant correlation
between the degree of interaction and the observed in vivo
neuroleptic effect, measured in the metrazole electro-shock
test (MES) was observed as shown in Fig. 117).
log I/ISo m~= 0.722 (0.12) log ln2+ 3.00 (0.19)
(2a)
n = 13 ? = 0.94 s = 0.086 ~ c , , , v a l . = 0.92 F = 178
The increase in interaction was paralleled by an increase
in the pharmacological effect over a large range of activity.
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
log
Flgure 11. Relationship between anticonvulsive activity, log l/Iw M=. and
strength of interaction of flupirtin derivatives with phospholipid vesicles
expressed as change in NMR relaxation rate,
log In,,,,=0.722(0.12)log
lR2+3.00(0.19)
n = 13
r = 0.97
r2 = 0.94 s = 0.086
= 0.92
F = 176
(taken from 7))
Inz
It is also obvious that the correlation obtained using octanol/water partition coefficients of the compounds instead of
is inferithe degree of interaction with phospholipids ( ln2)
or (?c,v, = 0.678)’).
log 1/Iso M s = 0.41 1 (0.134) log k’w,,+ 3.482 (0.23) (2b)
n = 13 1.2 = 0.79 s = 0.158 ~,,,,, = 0.678 F = 45
Especially those derivatives bearing a polar substituent at
the benzene ring deviate from the regression. This indicates
again that the octanol/water partition coefficient is not a
sufficient system if other than bulk effects are involved.
A change in lipophilicity at the ethyl carbamate group of
the pyridine ring leads to a decrease in activity. Probably
the distribution of hydrophobic surface with respect to the
cationic center of the flupirtin molecule is of importance. It
could influence the orientation of the drug molecule in the
membrane.
Another example where the biological effect could be
related to the partitioning into membranes is taken from a
paper of Choi and Rogers17).These authors have studied the
partition behaviour of adrenergic compounds using differently charged liposomes and also octanol. They found an
excellent linear correlation between the hypotensive effect
(PC2.3) and the partition coefficient (km2) determined with
neutral and negatively charged liposomes but not with log P
octanol (r = 0.69).
PC2.3 = -2.562 (1.13) log km2+ 6.33
n = 5 ? = 0.94 s = 0.228
scmSs
= 0.84
(3)
F = 51.6
We have determined the retention time for these derivatives on a phosphatidylserine covered column and found the
same excellent correlation with the hypotensive effect.
PCpj = -2.449 (0.79) log k,ppC + 3.44 (0.79)
(4)
n = 6 ?=0.95
s=O.21 ~crossva,,=0.91F=81.4
Arch. Phorm. (Weinheim)327.601-610 (1994)
607
Drug Membrane Interaction
An example where the inflammatory effect is correlated
with the broadening of the transition peak of DPPC vesicles
in the presence of a sulindac metabolite has also been
publishedI8).Unlike the active sulfite metabolite - produced
in vivo - neither sulindac nor the sulfone derivative show
any effect on the sharpness of the transition peak and are
pharmacological inactive. The change in cooperativity of
the lipid chains seems to be correlated with activity, i.e.
important for the functioning of antiinflammatoryagents.
Drug Membrane Interaction and Drug Resistance
Drug resistance is an increasing problem in the therapy of
diseases caused by bacteria as well as in tumor therapy.
Various mechanisms are responsible for the occurence of
drug resistance as for example:
- mutation or selection under therapy
- change in cell wall of target cells
- overproduction of target enzymes or proteins
- change in conformation of binding site.
Again we shall discuss two examples with respect to the
possible role of drug membrane interactions for resistance,
one on resistant bacterial cells, the other on resistant tumor
cells.
The first example is on E. coli, highly resistant towards
inhibitors of dihydrofolic acid reductase, DHFR, of the benzylpyrimidine type. The results are at the same time a warning to be careful in interpreting QSAR equations. From
QSAR analysis of a series of benzylpyrimidines tested
against drug sensitive and resistant E. coli Hansch and
coworkers concluded that one can design more effective
drugs against resistant bacteria and cancer cells by making
more lipophilic congeners19!
We have repeated such studies using resistant and sensitive E. coli and DHFR isolated from such bacteria. The
resistant strain used, E. coli RT 500, is a DHFR overproducer. In whole cell bacteria a steady decrease and in resistant cells an increase in inhibitory activity was noticed with
increasing chain length i.e. lipophilicity. So far this is in
agreement with the results of Hansch.
The results from the cell-free systems, using DHFR
extracted from those two E. coli strains show, however, that
there is no change in the conformation of the active site.
The inhibitory activity towards DHFR derived from sensitive and resistant cells follows the same ranking, trimethoprim being about 10 times more effective than the other
derivatives, no influence of chain length (Table 4) was
observed.
On the first glance this could support the hypothesis of
Hansch, that more lipophilic derivatives are more easily
transported through the cell wall of resistant cells. However, comparing the dose response curves, it became obvious
that there was a dramatic difference, indicating a change in
mechanism of action with chain length in case of resistant
cells (Fig. 12).
k/ko
0.6-
a.4-
0.2-
r*.
-.-
I
5
0
1
1
10
15
x
-----
-
20
25 800(
[~moV11
Figure 12. Effect of GH-306on generation rate, k, of E. coli RT 500
(TMP resistant) = generation rate of control culture. (-) normal shape of
the dose response curve for derivatives with 5 4 methylene groups
This was further supported by the observation that these
derivatives show no synergism with sulfonamides in case of
resistant strains and acting antagonistically in combination
with trimeth~prim~).
The structure-activity relationship
found for the sensitive strain can be described by eq. ( 5 )
and for the resistant strain by eq. (6) using the capacity factor log k, from HPLC experiments on an octanol covered
reversed phase column.
Table 4: Inhibitory activity (Im, uM) of 3-OCH3,4-alkoxy-benzylpyrimidinesin cell-free and whole cell systems of E. coli ATCC 11775 (TMP sensitive)
and E. coli RT 500 (TMP resistant) and the lipophilicity descriptor log k',
Im [uM]cell-free
E. coli
ATCC 11775
RT 500
IS0[uM] whole cells
E. coli
ATCC 11775
RT 500
log k',
GH30X
Trimethoprim, TMP
GH 01
GH 02
GH 03
GH 04
GH 05
GH 06
Brodimoprim
0.0018
0.0023
0.02
0.014
0.0I7
0.013
0.015
0.018
0.021
0.0011
0.015
0.018
0.022
0.0018
Arch. Pharm. (Weinheim) 327,601-610 (1994)
0.97
1.46
4.99
13.43
17.52
29.2
31.2
0.81
1147
1168
885
738.3
163
55.6
18.6
118.4
0.211
0.706
1.299
1.849
2.427
3.001
3.575
1.382
608
Seydel. Coats, Cordes, and Wiese
log 1/150E.colisells. = - 0.939 log k', + 0.121 (log k',)2
+ 296
n=7
r=0.99
s = 0.10
F = 107
log 1&0
E.coli res, = -
r=0.99
(5)
It
drug rdditlon
0.276 log k',
+ 1.164 log(0.051 k, + 1) - 3.02
n=9
counts hn L
s = 0.167
(6)
F = 83.9
One can conclude that these benzylpyrimidines are not any
longer DHFR inhibitors in case of the resistant strain but
inhibit bacterial growth by interacting with the membrane
of this resistant strain as a detergent, a cationic soap! The
resistant strain is not only an overproducer of DHFR,but
has also defects in its membrane, it has partially lost its
hydrophilic outer core so that these compounds can partition into lipid A and may form micelle structures.
Finally some of our latest preliminary results on a very
current topic are presented: the reversal of multi drug resistance in tumor and plasmodia1 cells by amphiphilic drugs.
This property was first reported for the Ca++-antagonist,
verapamil.
In several papers it is argued that these amphiphilic compounds prevent binding of antitumor and antimalarial drugs
to a P-glycoprotein which transports the drug out of the
cells thus preventing accumulation of the drug in the infected resistant cells20~21).
The P-glycoprotein is overproduced
in several resistant cell types. We remain unconvinced of
the general applicability of this hypothesis because:
1. Resistance of very different cells (tumor, bacteria, plasmodia) can be reversed and also in cases where no overproduction of P-glycoprotein is observed
2. Amphiphilic drugs with very different structures and
conformation can reverse multi drug resistance.
Among potential mechanisms under investigation in our
laboratories, the ability to reverse drug resistance by chang-
0 1 2 3 4 5 6 l h r
Figure 13. Reversal of tetracycline resistance of E. coli by desipramine
and maprotiline (data taken fromz2)).
es in membrane fluidity (permeability) remains most attractive.
Support for this mechanism can be achieved from results
obtained from studies on tetracycline resistant E. coli. This
resistance was effectively reversed upon addition of known
membrane active drugs: desipramine, maprotiline (Fig.
13)22).
Table 5: (data taken from *I))
Reversal of Chloroquino Resi8tance w i t h Antidepressant Drug8
(according to Bitonti at a1. Sciancm 242:1301(1988))
Doxepin
Xaprotilino
150
1.40
595
0.27
Varapamil
1300*
hasodona
>1345
2.07
.
>>1 00
Arch. Pharm. (Weinheim) 327,601-610(1994)
609
Drug Membrane Interaction
For a series of catamphiphilic drugs the interaction with
phospholipids was studied using NMR relaxation measurements (Table 4). It was found that the ranking in interaction
strength for various drugs was similar to the reported ranking in the concentration required to reverse chloroquine
resistance in plasmodia (Table 5).
An interesting example comes from recent reports comparing reversal of doxorubicin resistance in MCF-7 tumor
cells by tricyclic drugs related to chlor~promazine~~).
Particularly intriguing was the finding that trans-flupentixol (a
neuroleptic drug) was about three times more effective in
reversal of resistance as compared to the cis-isomer. This
suggests stereospecific interaction with some biomolecule.
Both flupentixol isomers were subjected to NMR-binding
studies with the result that trans-flupentixol binds more
than two times stronger to liposomes of lecithin than the
cis-isomer (Fig. 14)’q8).
In this respect it is interesting to note, that it has just been
shown by Aftab et
that trans-flupentixol is also about
3 times more potent in the inhibition of protein kinase C.
The letter enzyme needs phosphatidylserine for its activation and is responsible for the phosphorylation of the glycoprotein and its drug-pumping activity. As phosphorylation
is in equilibrium with dephosphorylation by phosphatase 1
and 2A, the inhibition of protein kinase C by the interaction
of catamphiphilic drugs with the activating phosphatidylserine could be the mechanism as proposed by us (Fig. 15).
It could explain why compounds possessing very deviating structure but similar physicochemical properties are
active in reversal of multi drug resistance.
Before some concluding remarks it is noteworthy to point
to the pharmacological effects by lipids itself as for instance
their platelet activation activity, the histamine release by
phosphatidylserine, the various effects of lysophosphatidylcholine and last not least the anticancer activity of several
lipid ethers.
Conclusion and Summary
1
0
0.05
0.10
0.15
0.20
0.25 ng lecithin
Figure 14. Relation between MDR ratio (*) for reversal of doxrubicin resistance by cis- und trans-flupentixol and the degree of interaction of the
two stereoisomers with phospholipids determined as change in proton relaxation rate as a function of Lecithin concentration (data taken from”’).
’
phosPhotidylsecine
I
Some aspects of drug membrane interaction and its influence on drug transport, accumulation, efficacy and resistance have been discussed. The interactions manifest themselves macroscopically in changes in the physical and thermodynamic properties of “pure membranes” or bilayers. As
various amounts of foreign molecules enter the membrane,
in particular the main gel to liquid crystalline phase transition can be dramatically changed. This may change permeability, cell-fusion, cell resistance and may also lead to
changes in conformation of the embedded receptor proteins.
Furthermore, specific interactions with lipids may lead to
drug accumulation in membranes and thus to much larger
concentrations at the active site than present in the surrounding water phase. The lipid environment may also lead
to changes in the preferred conformation of drug molecules.
These events are directly related to drug efficacy.
The determination of essential molecular criteria for the
interaction could be used to design new and more selective
therapeutics. This excursion in some aspects of drug membrane interaction underlines the importance of lipids and
their interaction with drug molecules for our understanding
of drug action, but this is not really a new thought but has
been formulated in 1884 by THUDICUM:
“Phospholipids are the centre, life and chemical soul of all
bioplasm whatsoever, that of plants as well as of animals”.
References
1
Figure 15. Hypothetical schematic drawing of the mechanism of action of
MDR reversing catamphiphiles
Arch. Pharm. (Weinheim) 327,601-610 (1994)
2
R.P. Mason, D.G.Rhodes, L.G. Herbette, J . Med. Chem. 1991, 34,
869-877.
J.K.Seydel, TiPS 1991.12.368-371.
610
J.K. Seydel, H.P. Cordes, M. Wiese, H. Chi. N. Croes, R. Hanpft, H.
Liillmann, K. Mohr, M. Patlen, Y. Padberg, R. Liillmann-Rauch, S.
Vellguth. W.R. Meindl, H. Schonenberger, Quonr.-Strue?.-Act.Relot.
1989,8,266-278.
E. Scheufler, T. Peters, Cell B i d Int. Rep. 1990.14, 381-388.
0. Jardetzky, G.C.K. Roberts, NMR in Molecular Biology, Academic
Press, 1981.
E.A. Coats, M. Wiese, H.-L. Chi, H.P. Cordes, J.K. Seydel, Quanf.
Sfruct.-Acf.Relor. 1992,lI. 364-369.
J.K. Seydel, M. Albores-Velasco, E.A. Coats, H.P. Cordes, B. Kunz,
M. Wiese, Quanf.Srrucf.-Acr.Relot. 1992,11,205-210.
J.K. Seydel, M. Wiese, H.P. Cordes, H.L. Chi, K.-J. Schaper, E.A.
Coats, B. Kunz, J. Engel, B. Kutscher, E. Emig in QSAR: Rationo1
Approaches fo the Design of Bioocrive Compounds. (Eds.: C. Silipo.
A. Vittoria), Elsevier Science Publishers. B.V., Amsterdam, 1991. pp.
361-376.
A. Koch, J.K. Seydel, A. Gasco, C. Tironi, R. Fruttero, Quonf.Struct.Act. Relot. 1993.12.373-382.
10 J.P. Akbaraly. J.J. Leng, G. Bozler, J.K. Seydel, in QSAR and Siroregies in !he Design of Bioocfive Compounds. (Ed.: J.K. Seydel). VCH,
Weinheim, 1985. p. 313.
11 R.C. Young, R.C. Mitchell, T.H. Brown, C.R. Ganellin, R. Griffiths,
M. Jones, K.K. Rana, D.Saunders. LR. Smith, N.E. Sore, T.J. Wilks,
J. Med. Chem. 19%8,31,656-67 1.
Seydel, Coats, Cordes, and Wiese
12
13
14
15
16
17
18
19
20
21
22
23
24
R. Schwyzer, Biochemistry 1986,25,4281-4286.
L.G. Herbette. Drug Design Delivery 1991,7,75-118.
H. Valdivia. R. Coronado, Biophys. J. 1988.53.555a.
E. Scheufler,J. Pharmocol. Exper. Ther. 1990,252,333-338.
L.G. Herbette, M. Trumbore, D.W.Chester, A.M. Katz, J . Mol. Cell.
Cardiol. 1988,20.373-378.
Y.W. Choi, J.A. Rogers,Phorm. Res. 1990,7,508-512.
Hwang San-Bao, T.Y. Shen, J . Med. Chem. 1981.24, 1202-121 I.
E.A. Coats, C.S. Genther, C. Dias Selassie, C.D. Strong, C. Hansch, J.
Med. Chem. 1985,28,1910-1916.
T. TSUNO,H. Iida, S. Tsukagoshi, Y. Sakurai, Cancer Research 1981,
41. 1967-1972.
S.K. Martin, A.M.J. Odoula, W.K. Milhous, Science 1987, 235, 899901.
J.K. Seydel, H.P. Cordes, M. Wiese, M. Kansy, B. Kunz in Trends in
MedicinoI Chemisrry, (Eds.: S . Sarel, R. Mechoulam, 1. Agranat).
Blackwell Scientific Publications, 1992, p. 397-401.
J.M. Ford, W.C. Prozialek, W.N. Hait, Mol. Pharmocol. 1989, 65.
105-115.
D. T. Aftab, W. N. Hait, Anal. Biochem. 1990,187, 84-88.
[Ph225]
Arch. Pharm. (Weinheim) 327,601-610 (1994)
Документ
Категория
Без категории
Просмотров
1
Размер файла
795 Кб
Теги
efficacy, drug, interactions, distributions, resistance, transport, importance, membranes, accumulation
1/--страниц
Пожаловаться на содержимое документа