Evaluation of substrate and inhibitor properties of a novel MDR modulator H17 towards transmembrane efflux pumps

Article abstract of DOI:10.1016/j.ejmech.2008.07.014

Substrate and inhibitor properties of H17 as novel modulator of transmembrane efflux pump activities have been characterized in an in situ absorption model. Poor substrate properties towards P-glycoprotein (P-gp) and multidrug resistance-associated protei

Full text of DOI:10.1016/j.ejmech.2008.07.014

European Journal of Medicinal Chemistry 44 (2009) 3060–3063
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Short communication
Evaluation of substrate and inhibitor properties of a novel MDR
modulator H17 towards transmembrane efflux pumps
,
Martin Richter b, Alexandra Richter b, Andreas Langner a, Andreas Hilgeroth a
*
a Institute of Pharmacy, Martin-Luther University Halle-Wittenberg, Wolfgang-Langenbeck Street 4, D-06120 Halle, Germany
b Baylor College of Medicine, Center for Cell and Gene Therapy, One Baylor Plaza, BCM 505 N1120 Houston, TX 77030, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Substrate and inhibitor properties of H17 as novel modulator of transmembrane efflux pump activities
have been characterized in an in situ absorption model. Poor substrate properties towards P-glycoprotein
(P-gp) and multidrug resistance-associated protein (MRP) have been demonstrated. In competition with
a MRP substrate H17 proved to have strong MRP-inhibiting properties. The profile of a strong inhibitor
with poor substrate properties makes H17 a perspective hopeful candidate for effective therapies of
transmembrane efflux pump activities.
Received 8 December 2006
Accepted 11 July 2008
Available online 19 July 2008
Keywords:
Absorption model
Efflux pump
Ó 2008 Elsevier Masson SAS. All rights reserved.
MDR modulator
Substrate properties
1. Introduction
drug use was strongly dose-limited due to toxic pharmacological
effects of the inhibitors [11]. In vivo evaluation of stronger
Transmembrane efflux pumps like P-glycoprotein (P-gp) and
the multidrug resistance-associated proteins (MRP) play an
important role in the resistance development against cytostatics
and antiretroviral therapeutics [1–4]. They transport drugs out of
the cells and thus lower intracellular therapeutically necessary
drug levels so that the P-gp or MRP expressing cell becomes
resistant [5]. An additional problem is the fact that the substrate
specificity of these efflux pumps is relatively low so that a drug-
resistant cell generally is found to be resistant also against other
possible therapeutics [5]. This low substrate specificity is dis-
cussed to result from several binding sites of the various
substrates to the transporter protein [6]. Thus, all the peptidic HIV
protease inhibitors used in antiretroviral HIV therapies show
cross-resistance as P-gp substrates [5].
Moreover, cells bearing the P-gp encoding gene like mdr1 are
induced to express more P-gp under therapy so that a P-gp over-
expression results in these cells [7]. This leads to the development
of viral sanctuaries in HIV therapy where P-gp is overexpressed at
the blood–brain barrier and the testes [8,9]. In the intestine the
expression of both efflux pumps P-gp and MRP leads to lowered
absorption rates of orally administered drugs [3,10].
inhibitors like valspodar showed dose-limited effects suggested
to result from the undesired inhibition of drug uptake trans-
porters [5]. Ritonavir as highly potent in vitro P-gp inhibitor also
disappointed in in vivo studies where only poor P-gp inhibiting
properties were found [7,12]. As far as investigated most P-gp
inhibitors are additionally substrates of the efflux pump so that
this property may also play a role in competitional uptake studies
of a substrate versus an inhibitor [13]. So the evaluation of both
inhibiting as well as substrate properties of a P-gp modulating
compound has to be considered in competition with relevant
compounds.
Recently, we reported H17 as novel non-peptidic HIV protease
inhibitor with P-gp inhibiting affinities in the range of verapamil in
an in vitro cell model [14]. To further characterize the new
compound and its potential to act as effective inhibitor we evalu-
ated its substrate properties towards transmembrane efflux pumps
in an in situ absorption model.
2. Chemistry
H17 has been synthesized as described [15]. Vinblastin has been
obtained from Gry-Pharma (Kirchzarten, Germany). Indomethacin
was a gift from Germed Berlin-Chemie (Berlin, Germany). Proben-
ecid was provided by Biokanol Pharma GmbH (Rastatt, Germany).
AE1 used as internal standard for the HPLC quantification of H17
was given by the acetylation of H17 in acetyl chloride under pyri-
dine base catalysis (Fig. 1).
Early inhibition of P-gp by drugs like cyclosporine or verap-
amil was proved in in vitro cell models, but in in vivo studies the
* Corresponding author. Tel.: þ49 345 55 25168; fax: þ49 345 55 27026.
E-mail address: andreas.hilgeroth@pharmazie.uni-halle.de (A. Hilgeroth).
0223-5234/$ – see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.07.014

M. Richter et al. / European Journal of Medicinal Chemistry 44 (2009) 3060–3063
3061
O
O
OH
H3C
O
CH3
OH
OH
O
N
N
H3CCOCl
pyridine
N
N
HO
O
O
O
O
CH3
H3C
H17
AE1
Fig. 1. H17 and the formation of the internal standard AE1.
3. Pharmacology
Similar effective permeabilities have been found for H17 alone
in all intestinal segments (Table 1). The combination of H17 with
vinblastine as used P-gp inhibitor leads to almost unchanged
absorption rates in the colon and the jejunal segment. Also in the
ileal segment the increase of H17 uptake is only very poor. For
talinolol as P-gp substrate increases of intestinal uptake rates of
more than 100% have been reported in combination with P-gp
inhibitor vinblastine [18]. So it is obvious that H17 practically is no
substrate of P-gp.
By the use of indomethacin as inhibitor of both MRP1 and MRP2
we find a poor increase of the intestinal uptake rate in the colon
segment with 24% while in the jejunal and the ileal segment the
increases of about 13% are negligible. In order to compare possible
increases of intestinal uptake rates of an MRP substrate in combi-
Experiments were performed in male White Wistar rats
purchased from Charles River Laboratories (Sulzfeld, Germany).
Rats were housed individually in a cage under controlled conditions
(21 ꢀC, 45% air humidity, 12-h light cycle). They were acclimatized
for at least two weeks before experiments and had access to tap
water and rodent pellet food from Altromin (Lage, Germany).
Animal studies were performed with approval from the Regier-
ungspra¨sidium Dessau/Sachsen-Anhalt.
Perfusion experiments were performed with weighted rats of
290–350 g. They were anesthesized with ketamine hydrochloride
(50 mg/kg) using Ketamine InresaÒ, Inresa Arzneimittel GmbH,
Freiburg, Germany, and with xylazin hydrochlorid, 10 mg/kg, in
RompunÒ 2% from Bayer, Leverkusen, Germany. Rats were placed
on a heating pad to maintain body temperature at 37 ꢀC. Then the
abdomen was opened by a midline longitudinal incision. In
nation with
a MRP inhibitor we furthermore investigated
indomethacin as MRP substrate itself and probenecid as strong
inhibitor of MRP [10]. While MRP1 and MRP2 have similar substrate
specificities not all of the inhibitors of MRP1 also inhibit MRP2 [10].
Probenecid as MRP1 inhibitor was also shown to inhibit MRP2 and
so was a suitable MRP inhibitor model compound. Indomethacin
showed poor effective permeabilities in the intestinal segments
(Table 2).
concentration dependent competition studies 30
combined with each inhibitor concentrations of vinblastine
(100 M) and 1 mM of indomethacin, respectively. In the indo-
methacin study 1 mM of the substrate was combined with 0.25 mM
of probenecid and 30 M of H17, respectively. Three intestinal
mM of H17 were
m
m
segements of each rat (n ¼ 3) were perfused over a length of each
7–10 cm (jejunum), w5 cm (ileum) and 2 cm (colon) in a three-step
procedure with a perfusate flow rate of 0.5 mL/min using an infu-
sion pump from Ismatec (Wertheim-Mondfeld, Germany). After
a preceding period of 30 min for reaching steady state with each
substrate and substrate/inhibitor containing tyrode buffer at
a temperature of 37 ꢀC, the outflow samples were taken over
30 min collecting in 5 min intervals. So for each perfusate and
intestinal segment six samples resulted.
However, lowest absorption rates have been observed in the
jejunal segment which is known to have high MRP1 expression
rates in rats [17]. The combination of indomethacin with proben-
ecid leads to an increase of the indomethacin uptake in all
segments. The highest increase of 70% has been found in the jejunal
segment due to the highest expression rates of MRP1 in this
segment. So compared to the much lower increased H17 uptake
rates reported above it can be concluded that H17 is also a poor
MRP substrate.
We used this model to additionally evaluate the MRP-inhibiting
properties of H17 in combination with the MRP substrate indo-
methacin. An increase in the effective permeabilities was found in
all segments. As has been found for probenecid the relatively
highest increase has been found in the jejunal segment with known
highest MRP expression rates. Although increases of substrate
uptake of 37–54% were lower than those for probenecid, they
nevertheless prove strong MRP-inhibiting properties for H17
because it was used in a lower competitor concentration than
probenecid due to solubility reasons.
4. Results and discussion
Within the intestine we investigated several rat gut segments by
perfusion: jejunum, ileum and colon. P-gp, MRP1 and MRP2 are
expressed within these intestinal segments [16,17]. While P-gp is
being located at the apical site of the epithelian cells and thus
lowers absorption rates from the intestine by the efflux of
substrates, MRP1 is located basolaterally and so tends to reduce
epithelial drug levels by the efflux into the body. MRP2 is also
located at the apical site [10].
In our perfusion studies we used vinblastine as proven P-gp
inhibitor at the given concentration which was reported to reach
high effects as P-gp inhibitor in the intestinal uptake of P-gp
specific substrates like talinolol [18]. As MRP inhibitor we used
indomethacin which has been demonstrated to have only MRP-
inhibiting properties without affecting P-gp [19].
Table 1
Effective permeabilities Peff [10ꢁ4 cm/s] for H17 alone and in combination with
vinblastin and indometacin, respectively
Intestinal segment
H17
H17 þ vinblastine
H17 þ indomethacin
Jejunum
Ileum
Colon
2.36 ꢂ 0.28
2.40 ꢂ 0.33
2.05 ꢂ 0.12
2.73 ꢂ 0.10
3.01 ꢂ 0.28
2.27 ꢂ 0.23
2.69 ꢂ 0.16
2.72 ꢂ 0.35
2.54 ꢂ 0.21

3062
M. Richter et al. / European Journal of Medicinal Chemistry 44 (2009) 3060–3063
Table 2
rate of 1 mL per minute resulting in retention times of 4 min for
H17 and 9 min for AE1. Detection was carried out UV-
spectroscopically at 240 nm wavelength. The specificity of the assay
was investigated by analyzing blank plasma samples. Precision and
accuracy were measured as inter- and intraassay precision in three
separate runs. Six replicates of five different concentrations of H17
(2500, 5000, 10,000, 20,000 and 30,000 ng/mL) were determined.
Analysis of variance (ANOVA) was used to calculate the inter- and
intraassay precision. The accuracy was calculated as the average
percentage of the nominal concentration. All analyses were within
10% at each concentration. The lower limit of quantification did not
exceed 10%. Samples were determined as six replicates. For the
limit of detection the difference between a spiked sample and
a background sample was tested with a paired t-test. Significance
was considered if p was less than 0.05. Quantities that gave a signal-
to-noise ratio of three were selected first for the determination of
the limit of detection. The value is 1000 ng/mL for the limit of
detection and 2500 ng/mL for the limit of quantification. Indo-
methacin perfusion samples were worked up and analyzed as
described in literature [20] using again the reverse phase chro-
matography column RP18 with a phosphate buffer (phosphoric acid,
potassium dihydrogen phosphate) pH ¼ 3.5 and methanol mixture
(42/58) as described at a flow rate of 1.5 mL per minute resulting in
a retention time of 2.4 min for indomethacin. Detection was carried
out UV-spectroscopically at 240 nm wavelength. All validation
procedures were realized as defined. The values for intra- and
interassay precision were within 10% at each concentration. The
lower limit of quantification for indomethacin was 180 ng/mL. The
value for the limit of detection was 50 ng/mL. Water transport was
calculated by weight measurement of each sample before and after
perfusion. Steady state existence regarding the analysts and water
uptake was examined via the trend test.
Effective permeabilities Peff [10ꢁ4 cm/s] for indomethacin alone and in combination
with probenecid and H17, respectively
Intestinal
segment
Indomethacin Indomethacin þ probenecid Indomethacin þ H17
Jejunum
Ileum
Colon
1.31 ꢂ 0.04
1.84 ꢂ 0.06
1.78 ꢂ 0.03
2.22 ꢂ 0.09
2.92 ꢂ 0.06
2.73 ꢂ 0.04
2.02 ꢂ 0.07
2.76 ꢂ 0.04
2.43 ꢂ 0.04
5. Conclusions
We demonstrated that H17 which belongs to a novel class of
non-peptidic HIV protease inhibitors shows poor substrate prop-
erties towards transmembrane efflux pumps compared to other
substrates discussed. As substrate properties are of great disad-
vantage with respect to clinical application as efflux pump inhibi-
tors these non-substrate properties of H17 prove this inhibitor to be
of favour compared to almost all reported inhibitors. Moreover,
within this single in situ model we additionally proved favourable
strong inhibitor properties towards efflux pumps. These inhibitor
properties give further perspectives for the application in oral
regimes with P-gp substrates which show poor oral absorption
rates in antiretroviral therapy like the HIV protease inhibitors do.
6. Experimental protocols
6.1. Chemistry
Commercial reagents were used without further purification.
The 1H NMR spectrum (400 MHz) was measured using tetrame-
thylsilane as internal standard. TLC was performed on E. Merck
5554 silica gel plates. The mass spectrum was measured with a MAT
710 mass spectrometer. Elemental analysis indicated by the
symbols of the elements was within ꢂ0.4% of the theoretical values
and was performed using a Leco CHNS-932 apparatus.
Water transport% ¼ 100$ðwin ꢁ woutÞ=win
Where win and wout represent the solution entering and exiting the
intestinal segment. The ratio wout/win was used to perform given
corrections concerning water transport.
6.1.1. 3,9-Dibenzyl-6,12-diphenyl-3,9-diazahexacyclo[6.4.0.02.7.04.11
.05.10]dodecane-1,5,7,11-tetrakis(methyleneacetate) (AE1)
H17 (0.009 g, 0.016 mmol) was dissolved in acetyl chloride
(0.126 g, 1.6 mmol) under stirring. Catalytic amounts of dried
pyridine were added to the mixture and stirring continued for 72 h
under TLC control. The reaction mixture was then acidified by the
addition of 10% of hydrochloride acid until a pH value of 4–5 was
reached.
7. Data analysis
With the determined mean concentrations of each P-gp
substrates in samples before perfusion (Cin) and after perfusion
(Cout) intestinal permeabilities (Peff) have been calculated as
previously described using the following equation [18]:
The water phase was extracted with dichlormethane for several
times. The organic layers were dried over sodium sulfate and after
filtration the solution volume was reduced in vacuum until the
compound crystallized (0.0062 g, 50%). White powder (m.p.:
Peff ¼ Q*ððCin=CoutÞ ꢁ 1Þ=2
p
rl
where Q* represents the flow rate, r means the radius and l the
length of the perfused intestine segment. From all single values,
arithmetical means and standard deviations (SD) have been
calculated. Statistical significance of correlations was determined
by performing linear regression analyses with tested significant
correlation coefficients p < 0.05. All studies were performed at least
in triplicates. Statistical significance was expressed using an
unpaired student’s t-test (INSTAT, V3.0).
199–202 ꢀC). 1H NMR (CDCl3):
d
¼ 2.04 (s, 12H, CH3CO), 3.12 (s, 4H,
H-2-, -4-, -8-, -10), 3.63 (s, 2H, H-6, -12), 3.71 (AB, CHBO,
2J ¼ 11.2 Hz), 3.96 (AB, CHAO, 2J ¼ 11.2 Hz), 4.24 (s, 4H, NCH2),
7.07–7.81 (m, 20H, aromatic H). ESIMS: m/z: 781 [M þ Hþ]. Anal.
C48H48N2O8 (C, H, N).
6.2. HPLC-assay
Acknowledgement
6.2.1. Sample preparation and analysis
Perfusion samples of H17 (500
mL) were spiked with AE1 as
The work was financially supported by the DFG.
internal standard (50 M) before extraction with chloroform (3 mL)
m
on a horizontal shaker. The centrifuged organic layer was removed
under nitrogen atmosphere at 38 ꢀC and the dried residue was
resolved in acetonitril/water (80/20) for HPLC analysis. Analysis of
samples before and after perfusion was carried out isocratically on
References
[1] W.T. Bellamy, Annu. Rev. Pharmacol. Toxicol. 36 (1996) 161–183.
[2] S.V. Ambudkar, S. Dey, C.A. Hrycyna, M. Ramachandra, I. Pastan,
M.M. Gottesmann, Annu. Rev. Pharmacol. Toxicol. 39 (1999) 361–398.
[3] R.B. Kim, M.F. Fromm, C. Wandel, B. Leake, A.J.J. Wood, D.M. Roden, J. Clin.
Invest. 101 (1998) 289–294.
a
reverse phase chromatography column RP18 (5
mm,
15 cm ꢃ 0.5 cm) with the eluent acetonitril/water (80/20) at a flow

M. Richter et al. / European Journal of Medicinal Chemistry 44 (2009) 3060–3063
3063
[4] J.D. Schuetz, M.C. Connelly, D. Sun, S.G. Paibir, P.M. Flynn, R.V. Srinavas,
A. Kumar, A. Fridland, Nat. Med. 5 (1999) 1048–1051.
[12] J. Drewe, H. Gutmann, G. Fricker, M. To¨ro¨k, C. Beglinger, J. Huwyler, Biochem.
Pharmacol. 57 (1999) 1147–1152.
[5] E.F. Choo, B. Leake, C. Wandel, H. Imamura, A.J.J. Wood, G.R. Wilkinson,
R.B. Kim, Drug Metab. Dispos. 28 (2000) 655–660.
[6] E.J. Wang, C.N. Casciano, R.P. Clement, W.W. Johnson, Biochem. Biophys. Acta
Protein Struct. Mol. Enzymol. 1481 (2000) 63–74.
[13] R.B. Kim, Top. HIV Med. 11 (2003) 136–139.
[14] A. Hilgeroth, C. Dressler, S. Neuhoff, P. Langguth, H. Spahn-Langguth, Phar-
mazie 55 (2000) 784–785.
[15] A. Hilgeroth, Mini Rev. Med. Chem. 2 (2002) 235–245.
´
´
[7] M.D. Perloff, L.L. Von Moltke, J.E. Marchand, D.J. Greenblatt, J. Pharm. Sci. 90
(2001) 1829–1837.
[16] M. Richter, N. Gyemant, J. Molnar, A. Hilgeroth, Pharm. Res. 21 (2004)
1862–1866.
[8] F. Thiebaut, T. Tsuruo, H. Hamada, M.M. Gottesmann, I. Pastan,
M.C. Willingham, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 7735–7738.
[9] C. Cordon-Cardo, J.P. O’Brien, D. Casals, L. Rittmann-Grauer, J.L. Biedler,
M.R. Melamed, J.R. Bertino, Natl. Acad. Sci. U.S.A. 86 (1989) 695–698.
[10] P. Borst, R. Evers, M. Kool, J. Wijnholds, J. Natl. Cancer Inst. 92 (2000)
1295–1302.
[17] S.D. Flanagan, C.L. Cummins, M. Susanto, X. Lui, L.H. Takahashi, L.Z. Benet,
Pharmacology 64 (2002) 126–134.
[18] A. Hanafy, P. Langguth, H. Spahn-Langguth, Eur. J. Pharm. Sci. 12 (2001)
405–415.
[19] M.P. Draper, R.L. Martell, S.B. Levy, Br. J. Cancer 75 (1997) 810–815.
[20] R.J. Stubbs, M.S. Schwartz, R. Chiou, L.A. Entwistle, W.F. Bayne, J. Chromatogr.
383 (1986) 432–437.
[11] R. Krishna, L.D. Mayer, Eur. J. Pharm. Sci. 11 (2000) 265–283.