Chalcogenopyrylium dyes as inhibitors/modulators of P-glycoprotein in multidrug-resistant cells

Article abstract of DOI:10.1016/j.bmc.2008.09.065

A series of chalcogenopyrylium dyes were evaluated as modulators/inhibitors of P-glycoprotein (Pgp). Their ability to inhibit verapamil (VER)-dependent ATPase activity (IC₅₀ values) in lipid-activated, mouse Cys-less mdr3 Pgp was determined. Their ability to promote calcein-AM (CAM) uptake in MDCKII-MDR1 cells and their capacity to be transported by Pgp in monolayers of MDCKII-MDR1 cells were also evaluated. The chalcogenopyrylium dyes promoted CAM uptake with values of EC₅₀ between 5 × 10^-6 and 3.5 × 10^-5 M and 7 of the 9 dyes examined in transport studies were substrates for Pgp with efflux ratios (P_BA/AB) between 14 and 390. Binding of three compounds (1-S, 3-S, and 4-S) to Pgp was also assessed by fluorescence. These three thiopyrylium dyes showed increased fluorescence upon binding to Pgp, giving apparent binding constants, K_app, on the order of 10^-7 to 10^-6 M. Compound 8-Te was particularly intriguing since it appeared to influence Pgp at low micromolar concentrations as evidenced by its influence on VER-stimulated ATPase activity (IC₅₀ of 1.2 × 10^-6 M), CAM uptake (EC₅₀ of 5.4 × 10^-6 M), as well as [³H]-vinblastine transport by Pgp in cells (IC₅₀ of 4.3 × 10^-6 M) and within inside-out membrane vesicles (IC₅₀ of 9.6 × 10^-6 M). Yet, Pgp did not influence the distribution of 8-Te in MDCKII-MDR1 monolayers suggesting that 8-Te may bind to an allosteric site.

Full text of DOI:10.1016/j.bmc.2008.09.065

Bioorganic & Medicinal Chemistry 16 (2008) 9745–9756
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Chalcogenopyrylium dyes as inhibitors/modulators of P-glycoprotein
in multidrug-resistant cells
Geri A. Sawada a, Thomas J. Raub a, J. William Higgins a, Nancy K. Brennan b, Teiah M. Moore b,
Gregory Tombline b,c, Michael R. Detty b,
*
a Drug Disposition, Eli Lilly and Company, Indianapolis, IN 46285, USA
b Department of Chemistry, University at Buffalo, The State University of New York, Buffalo, NY 14260-3000, USA
c Department of Microbiology and Immunology, University of Rochester Medical Center, 601 Elmwood Avenue, P.O. Box 607, Rochester, NY 14642, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of chalcogenopyrylium dyes were evaluated as modulators/inhibitors of P-glycoprotein (Pgp).
Their ability to inhibit verapamil (VER)-dependent ATPase activity (IC50 values) in lipid-activated, mouse
Cys-less mdr3 Pgp was determined. Their ability to promote calcein-AM (CAM) uptake in MDCKII-MDR1
cells and their capacity to be transported by Pgp in monolayers of MDCKII-MDR1 cells were also evalu-
ated. The chalcogenopyrylium dyes promoted CAM uptake with values of EC50 between 5 ꢀ 10ꢁ6 and
3.5 ꢀ 10ꢁ5 M and 7 of the 9 dyes examined in transport studies were substrates for Pgp with efflux ratios
(PBA/AB) between 14 and 390. Binding of three compounds (1-S, 3-S, and 4-S) to Pgp was also assessed by
fluorescence. These three thiopyrylium dyes showed increased fluorescence upon binding to Pgp, giving
apparent binding constants, Kapp, on the order of 10ꢁ7 to 10ꢁ6 M. Compound 8-Te was particularly
intriguing since it appeared to influence Pgp at low micromolar concentrations as evidenced by its influ-
ence on VER-stimulated ATPase activity (IC50 of 1.2 ꢀ 10ꢁ6 M), CAM uptake (EC50 of 5.4 ꢀ 10ꢁ6 M), as
well as [3H]-vinblastine transport by Pgp in cells (IC50 of 4.3 ꢀ 10ꢁ6 M) and within inside-out membrane
vesicles (IC50 of 9.6 ꢀ 10ꢁ6 M). Yet, Pgp did not influence the distribution of 8-Te in MDCKII-MDR1 mon-
olayers suggesting that 8-Te may bind to an allosteric site.
Received 20 August 2008
Revised 24 September 2008
Accepted 26 September 2008
Available online 30 September 2008
Keywords:
Chalcogenopyrylium dyes
P-glycoprotein
Multidrug resistance
Inhibitors
Transporters
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
Other important members of this family responsible for MDR in
cancer include the multidrug resistance protein series (including
The effective treatment of cancer is often thwarted by the
appearance of multidrug resistance (MDR) in patients.1–3 There
are three major mechanisms of drug resistance in cells including
decreased uptake of water-soluble drugs that require transporters
to enter cells; cellular changes that affect cell cycles, repair of DNA
damage, apoptosis, and drug metabolism; and increased efflux by
multidrug efflux pumps of hydrophobic drugs that might easily en-
ter the cell via diffusion through the plasma membrane.1 Multi-
drug efflux pumps are perhaps the most important contributor to
MDR and are capable of expelling multiple chemically and struc-
turally dissimilar drugs from the cell. The expression of the efflux
pumps can be up-regulated by exposure to a single chemothera-
peutic agent. The mammalian plasma membrane protein P-glyco-
protein (Pgp also known as MDR1 or ABCB1) is the most studied
MDR efflux protein. Pgp belongs to a larger protein family known
as the ATP Binding Cassette (ABC) proteins, the majority of which
are transporters.1 In humans, 48 members of this family are di-
vided into seven subclasses based upon degree of homology.2
multidrug resistance protein 1, MRP1 or ABCC1), breast cancer
resistance protein (BCRP, MXR, or ABCG2), and lung-resistance re-
lated protein (LRP).4
Complicating the treatment of MDR is the fact that Pgp is nor-
mally expressed at many physiological barriers including the
blood-brain barrier, the apical membranes of the epithelia, the
luminal surfaces of colon and small intestine, and kidney proximal
tubules.5 Numerous MDR reversal agents, such as verapamil (VER),
cyclosporin A, and PSC833 have been examined to minimize the ef-
fects of the various efflux pumps.1,3,6 However, these compounds
have major drawbacks, such as alterations in cell metabolism as
well as their toxicity toward normal tissues through interactions
with normally expressed Pgp. The therapeutic window for these
compounds is also severely restricted because the dose necessary
for effective inhibition of Pgp, in many cases, exceeds the minimal
toxic concentration in normal tissue.1,7,8 Furthermore, combina-
tions of the modulator and anti-cancer drugs have resulted in
unacceptable toxicity, which in turn has caused dose reductions
and less effective treatment.8 Thus, the search remains open for a
Pgp modulator/inhibitor that, when administered in combination
with chemotherapeutic agent(s), increases the anti-cancer drug
uptake, retention and effectiveness.
* Corresponding author. Tel.: +1 716 645 6800x2200; fax: +1 716 645 6963.
E-mail address: mdetty@buffalo.edu (M.R. Detty).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.09.065

9746
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
The rhodamines are an interesting class of molecules with re-
2. Results
spect to their interactions with Pgp. Many drugs have been as-
sayed for competitive transport with rhodamines in the NCI
screen.9 Tetramethylrosamine has been described as the best
transport substrate for Pgp both in viable MDR cells and in recon-
stituted Pgp.10 Computational studies have suggested that rhod-
amines and VER display compatible features in a rhodamine
pharmacophore model.11 Furthermore, rhodamine derivatives
provide a three-dimensional scaffold that can be incorporated
into either three-point or four-point pharmacophore models for
Pgp.12
We have recently examined a small library of tetramethylros-
amine derivatives for their ability to stimulate ATPase activity in
lipid-activated Cys-less mouse mdr3 Pgp.13 Within this library,
the substituents at the 9-position were varied systematically as
well as the chalcogen atom (O, S, Se) within the tricyclic xanthyli-
um core. What was intriguing within this series is that very slight
changes in structure gave up to 1000-fold differences in ability to
stimulate ATPase activity. Based on these results, we were able
to develop a rhodamine-based photosensitizer for drug-resistant
CR1R12 cells that over express Pgp that did not require an added
inhibitor or modulator to be effective.14
2.1. Selection and synthesis of chalcogenopyrylium compounds
We selected the 17 chalcogenopyrylium dyes shown in Chart 2
from the library of more than 150 similar structures that we have
generated over the years. The examples were chosen to resemble
the rhodamines and tetramethylrosamines in that at least two
amino- or dialkylaminophenyl substituents were present at the
2-, 4-, or 6-positions (Chart 1).16 The third position, R”, among
the 2-, 4-, and 6-substituents shown in Chart 1 was substituted
with alkyl, heteroaryl, and aryl substituents. The chalcogen atoms
were selected from S, Se, and Te. Dyes 9-S, 11-S, and 12-Se were
prepared for this study, as shown in Scheme 1, from chalcogeno-
pyranones 15-S, 16-S, or 16-Se.16c Values of logP for this series of
dyes were in the range 0.0–2.4 suggesting that the dyes were avail-
able to both aqueous and hydrophobic environments (Table 1).
2.2. Pgp ATPase activity
The chalcogenopyrylium compounds of this study displayed
variable ATPase stimulation of lipid-activated, mouse MDR3 Cys-
less Pgp, which is 87% identical to human MDR1 Pgp in sequence.17
ATP hydrolysis was determined via a spectrophotometric-coupled
assay. Stock solutions of VER and the chalcogenopyrylium dyes
were prepared in DMSO. For several of the chalcogenopyrylium
dyes in Chart 2 (1-S–4-S, 7-S, 8-S, and 8-Te) and for VER, a specific
activity was measured at a ‘‘saturating” concentration of drug, that
is, at a concentration where ATPase activity was unchanging (typ-
ically 5 ꢀ 10ꢁ5 to 1 ꢀ 10ꢁ6 M). Specific activities are compiled in
Table 1. Basal activity in the presence of DMSO alone (2% by vol-
ume, control samples performed in parallel) was 0.48 nmol/min/
As shown in Chart 1, the tricyclic xanthylium core is a rigid
pharmacophore fragment that will lock in the spatial orientation
of hydrogen bond acceptors and aromatic ring centroids and pro-
jection points. Diversity in structural flexibility can only come
from substituents in the 9-position. The 2,4,6-trisubstituted chal-
cogenopyrylium compounds shown in Chart
1 offer similar
structural features to the chalcogenorhodamines, but the aro-
matic centroids and projection points have rotational flexibility
and there are several sites for the introduction of structural
diversity.
l
g. Control VER-stimulated ATPase activity (2 ꢀ 10ꢁ4 M) for the
Earlier studies of the chalcogenopyrylium dyes with implanted
brain tumors in rats suggested that the chalcogenopyrylium com-
pounds passed the blood–brain barrier.15 While one reason might
be that the blood–brain barrier was compromised by the implan-
tation, an alternative explanation might be that the chalcogeno-
pyrylium photosensitizers were not exported by Pgp and
perhaps functioned to inhibit/modulate Pgp. In order to assess
whether the chalcogenopyrylium compounds are a new class of
Pgp inhibitors/modulators, we have examined a series of these
dyes for their ability to inhibit VER-induced ATPase activity in li-
pid-activated Cys-less mouse mdr3 Pgp, for their ability to pro-
mote calcein-AM (CAM) uptake into MDCKII-MDR1 cells, and
for their ability to be transported by Pgp in monolayers of
MDCKII-MDR1 cells. While several structural variations were able
to inhibit VER-induced ATPase activity and to promote CAM up-
take into MDCKII-MDR1 cells, one telluropyrylium derivative
showed minimal stimulation of ATPase activity in addition to
inhibiting VER-induced ATPase activity in lipid-bound Pgp, pro-
moting CAM uptake into and inhibition of [3H]-vinblastine efflux
by MDCKII-MDR1 cells, and inhibition of [3H]-vinblastine uptake
into Pgp inside-out membrane vesicles.
protein preparation used in this study was roughly 5.4-fold higher
at 2.6 nmol/min/
conferred by compound 3-S was 1.38 nmol/min/lg. Dyes 1-S, 8-S,
l
g. At 1 ꢀ 10ꢁ4 M 3-S, the specific ATPase activity
and 8-Te conferred specific activities that appeared nearly equiva-
lent to or below basal levels (0.3–0.48 nmol/min/ g) while dyes 2-
S, 4-S, and 7-S conferred intermediate specific activities in the
0.70–0.93 nmol/min/ g range (VER > 2-S, 4-S, 7-S > 1-S, 8-S, 8-Te).
l
l
2.3. Binding of chalcogenopyrylium dyes to Pgp as measured by
enhanced fluorescence
While the lack of stimulation of ATPase activity in certain of the
chalcogenopyrylium dyes may indicate that these compounds do
not interact with Pgp, lack of stimulation of ATPase activity accom-
panied by strong binding to Pgp is not unprecedented. We have de-
scribed a subset of rhodamine derivatives where apparent ATPase
activities are nearly equivalent to basal values.18 Values of KM, the
Michaelis–Menton constant, for drug stimulation of ATPase could
not be accurately measured in this set, yet these same rhodamine
derivatives promoted ATP occlusion at relatively low concentra-
Chart 1.

G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
9747
Chart 2. Chalcogenopyrylium dyes of this study.
Scheme 1. Preparation of chalcogenopyrylium dyes 9-S, 11-S, and 12-Se.
tions (ꢂ4 ꢀ 10ꢁ7 to 4 ꢀ 10ꢁ5 M). These data bolster the argument
that upon drug binding, the pump partitions to a drug-dependent
pathway and basal activity approaches zero.19 Further, this parti-
tioning phenomenon helps to explain cases where drugs appear
to confer ATPase stimulation near or even below basal values, yet
they are actually transported by Pgp. In those cases, it is argued
that any value above zero is attributed to the weak stimulation
by the drug that only appears similar to basal. Moreover, in addi-
tion to promotion of ATP occlusion, many rhodamine analogues
that failed to confer robust turnover were capable of inhibiting
VER-dependent ATPase activity at a similar concentration required
for ATP occlusion. This provided further evidence for recognition of
the analogues by Pgp; however, in these cases a direct measure of
binding was not presented.
We have recently shown that the relative fluorescence yields
from thiopyrylium dyes 1-S, 3-S, and 4-S increase dramatically
(>60-fold) upon binding to DNA relative to solution values.20 Should
the same phenomenon be observed upon binding of these dyes to
Pgp, then the enhanced fluorescence would provide a direct indica-
tor of binding to Pgp. Other compounds, including H33342 and LDS-
751, show greatly enhanced fluorescence emission upon binding to
the drug-binding pocket of Pgp, and this phenomenon has been used
to quantify directly drug interaction with purified Pgp.21
In order to provide a direct measure of binding to Pgp, we
attempted to assay the binding of thiopyrylium dyes 1-S, 3-S,
and 4-S using fluorescence. The advantage here compared to
the rhodamine series is that, similar to LDS-751, the thiopyryli-
um dyes display a very weak fluorescence in solution when
unassociated with a biopolymer, but display a large increase
in quantum yield of fluorescence upon binding.20 As illustrated
in Figure 1 for 4-S, increasing concentrations of the dyes 1-S, 3-
S, and 4-S does not lead to an increase in fluorescence in the
absence of Pgp in the presence of 0.2% Escherichia coli lipids
(‘‘sham” lipids). However, in the presence of 2.5 ꢀ 10ꢁ7 M Pgp
and 0.2% E. coli lipids, an increase in fluorescence is observed
with increasing concentrations of dye, which can be fit to give
values for an apparent binding constant, Kapp = 4.7 0.2 ꢀ 10ꢁ7
;
R2 = 0.99 for 4-S (Fig. 1). Similar values were found for 1-S
(Kapp = 4.5 2.9 ꢀ 10ꢁ6 M; R2 = 0.95) and for 3-S (Kapp = 2.2
1.2 ꢀ 10ꢁ6 M; R2 = 0.97; data not shown). The increase in fluo-
rescence is similar to that observed for binding of dyes 1-S,
3-S, and 4-S to DNA, with emission maxima at 690 nm.20 These
results suggest that the pyrylium dyes can bind to Pgp with
micromolar affinities (or lower, e.g., 4-S) without necessarily
conferring robust stimulation of ATPase activity (e.g., the near
basal activity found with 1-S).

9748
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
Table 1
Values of the n-octanol/water partition coefficient (log P), IC50 for the inhibition of VER-induced (2 ꢀ 10ꢁ4 M) and residual ATPase activity in MDR3 cys-less Pgp, specific activity
at a ‘‘saturating” dye concentration, EC50 for the uptake of CAM in MDCKII-MDR1 cells, and % inhibition relative to fully-inhibited MDCKII-MDR1 cells at 5 ꢀ 10ꢁ6 and
2.5 ꢀ 10ꢁ5 M chalcogenopyrylium dye
Compound
LogP
IC50, ꢀ 10ꢁ6
M
Residual ATPase activity
Specific activity
(nmol/min/lg)
CAM uptake EC50, ꢀ 10ꢁ6
M
5
lM, % LSN335984
25
lM, % LSN335984
(nmol/min/
lg)
None
VER
1-S
2-S
3-S
4-S
4-Se
5-S
5-Se
6-Se
7-S
8-S
8-Te
9-S
10-S
11-S
12-Se
13-Te
14-S
—
—
—
15
0.4 0.1
2.6 0.1
0.5 0.1
1.3 0.1
2.5 0.1
1.4 0.1
—
0.48 0.10
2.60 0.10
0.48 0.10
0.74 0.10
1.37 0.10
0.70 0.10
—
—
—
—
—
63
74
5.1
0.6
0.1
1.2
1.5
1.6
0.0
0.8
2.2
2.2
2.1
2.4
2.1
2.1
1.3
2.3
2.2
1.2
5.1 1.2
16
—
9.8 1.1
7.1 1.1
6.7 1.3
14
20
25
35
35 0.65
13
—
8
2
1
2
1.0
>100
0.88
—
5.3
—
>100
0.7
71
1.2
30
11
3.1
55
7.9
5.7
—
15.0 0.1
87
60
67
14
22
29
16
—
230
—
—
—
—
—
—
2
1
1
2
3
2
1
25
23
7
5.0 0.1
6.0 0.2
2.0 0.3
—
1
2
1
0.9 0.1
—
1
1
1
1
—
—
2.5 0.1
1.6 0.1
0.0 0.1
0.4 0.1
1.0 0.1
0.7 0.1
1.0 0.1
0.4 0.1
0.9 0.1
0.9 0.1
0.93 0.10
0.30 0.10
0.44 0.10
—
—
—
—
—
—
—
5.4 1.0
32
—
3
3
—
—
—
—
—
—
—
—
—
—
—
ATPase activity. We next examined the effect of increasing concen-
trations of the chalcogenopyrylium dyes on Pgp ATPase activity
stimulated by the presence of a fixed, saturating concentration of
VER (2 ꢀ 10ꢁ4 M) using lipid-activated, mouse MDR3 Cys-less
Pgp. Values of IC50, the concentration giving 50% reduction of the
VER-induced ATPase stimulation, and the residual ATPase activity
(nmol/min/lg), are given in Table 1. Typical plots are shown in
Figure 2 for 4-S and 8-Te.
This assay indicated that several of the chalcogenopyrylium
dyes, in fact, had values of IC50 near 10ꢁ6 M—2-S, 4-S, 5-S, 7-S, 8-
Te, 11-S, 13-Te, and 14-S (Fig. 2 for 4-S and 8-Te). Of the 17 dyes
examined, only two (3-S and 6-Se) failed to display an apparent
inhibitory effect with values of IC50 >10ꢁ4 M (Table 1). The majority
of the dyes examined had intermediate residual values of ATPase
activity as illustrated in Figure 2A for 4-S with residual ATPase
Figure 1. Relative fluorescence of 4-S in ET buffer and 0.2% E. coli lipids with (filled
circles) and without (open circles) 2.5 ꢀ 10ꢁ7 M Pgp. In the presence of Pgp, Kapp is
(4.7 0.2) ꢀ 10ꢁ7 M with R2 = 0.99.
activity of 1.4 nmol/min/lg. Near basal levels of ATPase activity
were also observed for the residual ATPase activity with 1-S, 8-
Te (Fig. 2B), and 12-Se indicating that in the case of these deriva-
tives, when the chalcogenopyrylium dye reaches its ‘‘saturation
point,” (i.e., higher concentrations would not lower residual ATPase
activity further) only very low levels of ATPase activity can be
measured.
2.4. Inhibition of VER-induced stimulation of ATPase activity
The fluorescence studies suggested that the chalcogenopyryli-
um dyes might be able to inhibit VER-induced stimulation of Pgp
Figure 2. Inhibition of VER-induced (2 ꢀ 10ꢁ4 M) stimulation of ATPase activity in lipid-activated mouse cys-less Pgp with A) thiopyrylium dye 4-S and B) telluropyrylium
dye 8-Te. Values of IC50 were determined by fitting the data to a decreasing slope with a single exponential parameter for the dose-response. Data points were the average of
three experiments which showed excellent agreement (<10% standard deviation at each point).

G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
9749
The residual ATPase activity remaining after inhibition of VER-
induced stimulation of ATPase activity may be due to either resid-
ual VER stimulation of ATPase or may be due to stimulation by the
chalcogenopyrylium dye. If purely competitive inhibition is occur-
ring, the residual ATPase activity should correspond to the residual
activity associated with the dye itself. While the residual ATPase
activity corresponds nicely to specific activity values for chalco-
genopyrylium dyes alone (1-S, 4-S, 7-S, 8-S, and 8-Te in Table 1),
the possibility of non-competitive inhibition or more complex
interactions cannot be ruled out.
In order to address competitive and non-competitive binding of
VER and the chalcogenopyrylium dyes, we next examined the
binding of VER and telluropyrylium dye 8-Te using classical com-
petitive analysis for ATPase inhibition. Competitive analysis has
been used by other labs to establish relationships between drugs
and Pgp as well as in our analysis of the interaction of rhodamines
with VER.18,22 The apparent competitive inhibition pattern of tell-
uropyrylium dye 8-Te with VER is shown inFigure 3. The data were
fit to a modified partition equation for competitive inhibition18,19
where basal ATPase activity is replaced by drug-induced stimula-
tion. In this particular preparation of Pgp, the basal ATPase activity
to other preparations, maximal VER-induced stimulation of ATPase
activity, Vmax, was 2.67 0.08 nmol/min/lg. Importantly, the KM
D
for VER stimulation (or the amount of VER required for 50% stim-
ulation), was measured to be (1.6 0.4) ꢀ 10ꢁ5 M, which is in
D
excellent agreement with reported KM values of 2 ꢀ 10ꢁ5 M.13
Based on these data, non-linear least-squares regression analysis
gives a half-maximal inhibition constant, Ki, of (4 1) ꢀ 10ꢁ8
M
for telluropyrylium dye 8-Te for inhibition of VER-induced stimu-
lation of Pgp ATPase activity.
Although the fits show slight deviation from the raw data at
lower VER concentrations, the fit appeared more consistent with
the data at higher VER concentrations. Essentially, while the cur-
rent data suggest competitive behavior at higher VER concentra-
tions, it also appears likely that more complex interactions
within the drug binding pocket (or an allosteric site) may exist.
Therefore, we hesitate to speculate further until it is possible to
perform more detailed experiments that may shed light on stoichi-
ometry and the location of the chalcogenopyrylium binding site(s).
Additional knowledge of stoichiometry or further competitive
analyses with other drugs may allow additional parameters to be
incorporated when fitting the data. It may also be notable that
comparison of the apparent competitive behavior of 8-Te pre-
sented herein with the behavior of the rhodamine derivatives ana-
lyzed previously also suggests that the interaction of the pyryliums
is distinct from the rhodamines, which more closely resembled a
strict competitive pattern.
was slightly elevated (0.83 0.03 nmol/min/lg); however, similar
2.5. Enhancement of CAM uptake into MDCKII-MDR1 cells
Selected chalcogenopyrylium compounds were next evaluated
for their ability to facilitate the uptake of CAM into MDCKII-
MDR1 transfected cells, which over express Pgp (or ABCB1).23 Sev-
eral of the chalcogenopyrylium compounds with values of IC50
near 1 ꢀ 10ꢁ6 M (4-S, 7-S, and 8-Te) and two of the chalcogeno-
pyrylium compounds with residual ATPase activity near basal val-
ues and values of IC50 <2 ꢀ 10ꢁ5 M (1-S and 5-S) were selected as
were compounds 3-S and 6-Se, which showed no apparent inhibi-
tion of VER-induced ATPase activity at concentrations up to 1 ꢀ 10
ꢁ4 M. Uptake of CAM was followed by measuring relative fluores-
cence values obtained after 20-min incubation with CAM at
37 °C. These values were plotted as a function of the concentration
of dye as shown in Figure 4 for dyes 4-S and 8-Te, and values of
EC50 of uptake of CAM are compiled in Table 1.
Figure 3. Classic competitive analysis for inhibition of VER-dependent ATPase by
analogue 5 was conducted as described in Experimental Procedures. The lines
represent the non-linear least-squares regression fits of the data to an equation
consistent with competitive inhibition which also takes into account the partition
model (see reference 18) and give a Ki of (4 1) ꢀ 10ꢁ8 M. Data represent the
average of two separate experiments which showed excellent agreement (a
standard deviation of <10% for all points).
Interestingly, all nine compounds facilitated the uptake of CAM
into MDCKII-MDR1 cells including 3-S (EC50 of 9.8 ꢀ 10ꢁ6 M) and
6-Se (EC50 of 2.5 ꢀ 10ꢁ5 M). Chalcogenopyrylium dyes 4-S, 4-Se,
Figure 4. Uptake of CAM into MDCKII-MDR1 cells as a function of concentration of (A) chalcogenopyrylium dye 4-S and (B) chalcogenopyrylium dye 8-Te. Values of EC50
were determined by a sigmoidal dose-response (variable slope) analysis.

9750
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
and 8-Te were the three most potent compounds in this assay with
values of EC50 of 7.1 ꢀ 10ꢁ6 M, 6.7 ꢀ 10ꢁ6 M, and 5.4 ꢀ 10ꢁ6 M,
respectively.
BA directions in the absence or presence of inhibitor are compiled
in Table 2. Of the nine compounds examined in the MDCKII-MDR1
cell monolayer system, seven of the compounds were shown to be
Pgp substrates as indicated by net efflux defined by BA/AB Papp ra-
tios of 14–390. Curiously, Compounds 7-S and 8-Te appeared to be
non-substrates as indicated by an efflux ratio of <2. What is most
intriguing, especially in the case of 8-Te, there appeared to be no
correlation between the magnitude of the efflux ratio in this assay
when compared to inhibition of ATPase or promotion of CAM up-
take. All nine chalcogenopyrylium dyes showed slow to very slow
passive diffusion when Pgp was inhibited by LSN335984 as defined
by a decrease in the efflux ratio to ꢂ1.0. It is unclear how the mea-
sured 10-fold faster passive Papp for compounds 3-S and 7-S, albeit
still slow or membrane limited, is in response to the n-propyl ver-
sus p-aminophenyl substitutions that were made. Another inter-
esting observation is that PAB for 3-S in the uninhibited assay is
more than an order of magnitude smaller than Papp in the fully
inhibited system (1.5 nm sꢁ1 vs. 22 nm sꢁ1, respectively).
Cell partitioning of the chalcogenopyrylium dyes appears to be
highly polarized in the AB direction in the cell monolayers, which
suggests that they may reside preferentially in the outer leaflet of
the membrane. When Pgp is inhibited with 2.5 ꢀ 10ꢁ6 M
LSN335984, the polarization disappears in seven of the nine com-
pounds and is reduced in the remaining two (4-S and 4-Se, Table
2). In all nine compounds, the % cell-associated dye in the AB direc-
tion is reduced in the presence of 2.5 ꢀ 10ꢁ6 M LSN335984 while
the % cell-associated dye in the BA direction is unaffected by
2.5 ꢀ 10ꢁ6 M LSN335984.
As a check of the IC50 values, the CAM uptake in MDCKII-MDR1
was repeated at concentrations of 5 ꢀ 10ꢁ6 M and 2.5 ꢀ 10ꢁ5 M in
the chalcogenopyrylium dye and the inhibition of Pgp was compared
as a percentage of the inhibition observed with of 2.5 ꢀ 10ꢁ6 M
LSN335984 (IC50 = 4 ꢀ 10ꢁ7 M), which completely inhibits Pgp.
LSN335984 is structurally related to the Pgp-specific inhibitor
LY335979 or zosuquidar.3 Values of the percentage inhibition at
the two concentrations are given in Table 1. Compounds 1-S, 3-S,
4-S, and 4-Se were all comparable in this assay with inhibition of
13–25% at 5 ꢀ 10ꢁ6 M and 60-87% at 2.5 ꢀ 10ꢁ5 M relative to
LSN335984 inhibition. These values are consistent with IC50 values
between 5 ꢀ 10ꢁ6 and 2.5 ꢀ 10ꢁ5 M, which is observed (Table 1).
Compound 8-Te gave 32% inhibition at 5 ꢀ 10ꢁ6 M and complete
inhibition (>100% relative to LSN335984) at 2.5 ꢀ 10ꢁ5 M, which
is consistent with an IC50 value near 5 ꢀ 10ꢁ6 M (observed IC50 of
5.4 ꢀ 10ꢁ6 M, Table 1). In contrast, dyes 6-Se and 7-S showed min-
imal inhibition at 5 ꢀ 10ꢁ6 M and 2.5 ꢀ 10ꢁ5 M.
2.6. Transport across monolayers of MDCKII-MDR1 cells
The transport of 5 ꢀ 10ꢁ6 M solutions of the same set of dyes
used to evaluate CAM uptake was examined in monolayers of
MDCKII-MDR1 transfected cells.23 MDCKII-MDR1 monolayers dis-
play apical and basolateral polarized membranes and are consid-
ered to be a near-physiological model for studying Pgp drug
transport. In the monolayer, Pgp is asymmetrically distributed
being solely present at the apical membrane. In order to evaluate
the role of Pgp in the transport of the compounds of this series,
transport was measured in the absorptive (apical to basolateral
or AB) and secretory (basolateral to apical or BA) transport direc-
tion of the cell monolayer, and then calculated an efflux ratio
(PBA/AB).24 BSA addition to the buffer was required because a
marked fraction of mass added to the donor equilibrated with
the cell monolayer for some of the compounds and this resulted
in gross underestimation of the permeability constant.25 The assay
was repeated in the presence of 2.5 ꢀ 10ꢁ6 M LSN335984 to mea-
sure transport when Pgp was fully inhibited. The ratio PBA/AB re-
ported in Table 2 is normalized—that is, by dividing PBA/AB in the
absence of inhibitor by the secretory/basal ratio observed in the
fully inhibited system. Large efflux ratios are assumed to be due
to high Pgp-mediated efflux of the compound. Values of passive
transport (Papp) in the presence of inhibitor, transport in the
absorptive (PAB) and secretory (PBA) mode in the absence of inhib-
itor, the normalized ratio [PBA/AB (no inhibitor)/PBA/AB (with inhibi-
tor)], and the % cell-associated chalcogenopyryliun dye in AB and
2.7. Inhibition of vinblastine transport in a membrane vesicle
model
Compound 8-Te was evaluated for its ability to inhibit vinblas-
tine transport by Pgp in a membrane vesicle model. The accumula-
tion of [3H]-vinblastine into Pgp membrane vesicles in the
presence of 8-Te (2.5 ꢀ 10ꢁ7 M to 2.5 ꢀ 10ꢁ5 M) was monitored,
using 5 ꢀ 10ꢁ6 M LSN335984 as a standard for complete Pgp inhi-
bition. Previously, the inhibition of [3H]-vinblastine accumulation
in Pgp membrane vesicles by LSN335984 gave an IC50 value of
(6.5 1.1) ꢀ 10ꢁ9 M and, thus, complete inhibition of Pgp was as-
sumed at 5 ꢀ 10ꢁ6 M. Inside-out membrane vesicles26,27 were pre-
pared from human embryonic kidney (PEAKSTABLE) cells over
expressing human Pgp. Mixtures of [3H]-vinblastine and 8-Te were
added to the vesicles and, following a reaction time of 2.5 min, the
vesicles were removed by filtration and the residual radiation on
the filter was measured by scintillation counting to give the
relative values shown in Figure 5A with an IC50 of (4.3 1.8) ꢀ
10ꢁ6 M and a slope of 1.8 1.0.
Table 2
Permeability in the MDCKII-MDR1 cell monolayer
Passive Papp nm sꢁ1
PAB, nm sꢁ1
PBA, nm sꢁ1
Efflux ratiob PBA/AB
Uninhibited % cellc (AB/BA)
Inhibited % cellc,d (AB/BA)
a
Compound
1-S
3-S
4-S
4-Se
5-S
5-Se
6-Se
7-S
4.2 0.7
5.8 0.7
1.5 0.7
5.3 0.5
2.3 0.2
3.7 0.2
3.5 0.1
3.4 0.1
51 11
370 76
340 38
14
390
134
161
70
57
17
1.6
1.4
(65 1)/(25 1)
(27 1)/(11 1)
(73 2)/(37 1)
(63 4)/(26 3)
(55 1)/(24 1)
(50 1)/(18 1)
(27 1)/(12 1)
(55 1)/(19 1)
(7 1)/(3 1)
(24 3)/(17 3)
(8 1)/(11 1)
(44 1)/(25 8)
(37 4)/(23 4)
(19 1)/(20 1)
(18 1)/(13 4)
(14 1)/(10 1)
(16 1)/(17 1)
(7 1)/(5 1)
22
3
3.4 1.1
2.8 0.7
3.1 0.6
2.8 0.5
2.8 0.7
32 0.7
4.8 0.7
290
170
120
45
6
1
3
2
53
4
42 29
4.2 0.2
8-Te
8.5 0.3
a
Apparent permeability coefficient with Pgp inhibited by 2.5 ꢀ 10ꢁ6 M LSN335984.
b
c
The ratio of Papp in the BA direction to Papp in the AB direction divided by the ratio of Papp in the BA direction to Papp in the AB direction in the fully inhibited system.
% Cell associated is the fraction of mass extracted from the cell monolayer by methanol wash after 1-h flux in the AB or BA direction.
Pgp inhibited by 2.5 ꢀ 10ꢁ6 M LSN335984.
d

G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
9751
since VER-induced stimulation of Pgp is completely inhibited
(residual ATPase activity of basal value or less) by several of the
dyes (1-S, 8-S, 8-Te, and 12-Se).
All nine chalcogenopyrylium dyes that were evaluated with
MDCKII-MDR1 cells, including two that showed no apparent inhi-
bition of VER-induced ATPase activity (3-S and 6-Se), were effec-
tive at promoting CAM uptake with EC50 values in the 5.4 ꢀ 10
ꢁ6 M to 2.5 ꢀ 10ꢁ5 M range. In particular, one telluropyrylium
dye, 8-Te showed a low IC50 (1.2 ꢀ 10ꢁ6 M) for inhibiting VER-in-
duced ATPase activity, an even lower Ki of 4 ꢀ 10ꢁ8 M for compet-
itive inhibition of VER-induced ATPase activity, a low level of
residual ATPase activity (0.4 nmol/min/lg) that appears equivalent
to basal values, and a low EC50 (5.4 ꢀ 10ꢁ6 M) for CAM uptake into
MDCKII-MDR1 cells.
The lack of inhibition of VER-induced ATPase activity with 3-S
was somewhat surprising to us since the fluorescence studies indi-
cated Kapp of (2.2 1.2) ꢀ 10ꢁ6 M for 3-S and since 3-S promotes
CAM uptake in MDCKII-MDR1 cells with an EC50 of 6.4 ꢀ 10ꢁ6 M.
Furthermore, at 2.5 ꢀ 10ꢁ5 M 3-S, 87% of the inhibition of
2.5 ꢀ 10ꢁ6 M LSN335984 (EC50 = 4 ꢀ 10ꢁ7 M), which completely
inhibits Pgp, was observed with respect to CAM uptake (Table 1).
The specific activity measured for 1 ꢀ 10ꢁ4 M 3-S with Pgp was
80% of the specific activity induced by 2 ꢀ 10ꢁ4 M VER (Table 1).
One possible explanation is that 3-S binds noncompetitively at
an allosteric site with respect to VER. A second possibility is that
VER and 3-S are exactly compensatory in their interaction with
Pgp. Finally, given its low % cell association in the absence of Pgp
(see below) differences between experimental conditions or Pgp
environment (i.e., membrane composition) may also account for
differences in behavior between ATPase and CAM experiments.
The possibility that chalcogenopyrylium dyes and VER can bind
simultaneously to Pgp has ample precedent. Shapiro and Ling28 de-
scribed two functional transport sites within Pgp: the R-site, which
interacts preferentially with rhodamine 123, and the H-site, which
interacts preferentially with H33342. Binding of a drug to one site
gives allosteric effects on the other. In their analysis, binding of a
drug to the H-site (or R-site) stimulates transport of an R-site (or,
conversely, H-site) drug, while inhibiting transport of other H-site
(or R-site) drugs. However, it is not clear whether these functional
sites have distinct physical locations within the protein. Lugo and
Sharom21b found that two compounds, rhodamine 123 and LDS-
751, which compete with each other for transport by Pgp, can both
bind to the R-site simultaneously. In this case, they appeared to
interact with each other non-competitively, rather than competi-
tively and the outcome was a 5-fold negative effect on each other’s
binding. It was suggested that the two drugs may bind to different
overlapping regions, or minipockets, within a large flexible binding
site.21b Loo, Bartlett, and Clarke have shown that a methanethiosul-
fonate derivative of rhodamine can covalently bind to human Pgp
in an activated state.29 However, ATPase activity of this rhoda-
mine-Pgp complex is further stimulated by VER and is inhibited
by Hoechst 33342, indicating that covalent attachment of the rho-
damine still permits VER or Hoechst 33342 to bind to Pgp, presum-
ably at a different site. These same authors have also shown that
specific mutations in Pgp affect the binding of specific subsets of
drugs again indicating that different binding sites are involved.30
In a comparison of thiopyrylium and selenopyrylium dyes, the
pairs 4-S/4-Se and 5-S/5-Se showed little difference in CAM uptake
as a function of heteroatom expressed as values of EC50 or as a per-
centage of LSN335984 in the two-dose inhibition screen (Table 1).
In contrast, telluropyrylium dye 8-Te was far more potent than its
thiopyrylium analogue 8-S at inhibiting VER-induced ATPase
activity with values of IC50 of 1.2 ꢀ 10ꢁ6 M and 7.1 ꢀ 10ꢁ5 M,
respectively. Irrespective of IC50 values, both 8-S and 8-Te had very
low values for residual ATPase activity when VER-induced ATPase
Figure 5. Analyses for inhibition of vinblastine transport by wild-type, human Pgp
using 8-Te were conducted as described in the experimental procedures. (A) Dose
response for membrane vesicles from PEAK-MDR1 cells. (B) Dose–response for
MDCKII-MDR1 cells. The lines represent the non-linear least-squares regression fits
of the data giving IC50 values of (4.3 1.8) and (9.6 0.8) ꢀ 10ꢁ6 M, respectively.
The slopes of the fits were 1.8 1.0 and 2.5 0.5, respectively. Data represent the
average and standard error mean for duplicate measurements. Two separate
experiments in different days gave similar results (data are from one of those
studies).
2.8. Inhibition of vinblastine efflux by 8-Te in MDCKII-MDR1
cells
[3H]-Vinblastine in an appropriate dilution series with 8-Te and
BSA was introduced to the basolateral chamber of a monolayer of
MDCKII-MDR1 transfected cells. The appearance of [3H]-vinblas-
tine in the apical chamber was monitored by scintillation counting
as shown in Figure 5B and gave an IC50 of (9.6 0.8) ꢀ 10ꢁ6 M with
a slope of 2.5 0.5. The vinblastine transport data from both cells
and prepared vesicles are also interesting because the inhibition
IC50 values do correlate with ATPase and CAM inhibition data.
3. Discussion
The goal of this study was to evaluate chalcogenopyrylium dyes
as potential Pgp substrates and/or modulators. Given the enhanced
flexibility and capacity for chemical variability compared to rhod-
amines, if tight-binding pyryliums could be uncovered, they could
serve as a novel scaffold to develop Pgp inhibitors. Empirically, sev-
eral structural variations of the chalcogenopyrylium dyes in Chart
2 were able to inhibit VER-induced ATPase activity and to promote
CAM uptake into MDCKII-MDR1 cells. Seven of the nine com-
pounds evaluated in MDCKII-MDR1 cell monolayers were identi-
fied as substrates for Pgp with efflux ratios (PBA/AB) of 14–390.
These dyes appear to be modulators/inhibitors of Pgp function
and indeed represent a new class of compounds with this function.
Binding of several of the dyes to Pgp could be demonstrated by
the enhanced fluorescence observed with increasing concentration
of dye (Fig. 1) suggesting values of Kapp on the order of 10ꢁ7 to 10
ꢁ6 M for 1-S, 3-S, and 4-S. While the fluorescence binding studies
for these three compounds do not indicate the stoichiometry of
binding, one can reasonably infer that most of the Pgp is bound
activity was fully inhibited (0.0 and 0.4 nmol/min/lg, respec-

9752
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
tively). Telluropyrylium dye 13-Te also displayed a relatively low
value of IC50 (7.9 ꢀ 10ꢁ6 M) for VER-induced ATPase activity and
tine transport in a membrane vesicle model (1.2 ꢀ 10ꢁ6 M and
4.3 ꢀ 10ꢁ6 M, respectively). If there are other transporters in the
cell membrane in the MDCKII-MDR1 cells, they have minimal im-
pact on the performance of 8-Te as an inhibitor of Pgp.
residual ATPase activity of 0.9 nmol/min/lg.
Mechanistically, the chalcogenopyrylium compounds pose
some interesting dilemmas. In isolated Pgp, these compounds have
varied effects on VER-induced stimulation ranging from no impact
at all (3-S, 6-Se) to very low turnover which appears as essentially
complete inhibition (1-S, 8-S, 8-Te, 12-Se). Of nine compounds
examined, seven are substrates for Pgp and two compounds are
not (7-S, 8-Te) as determined by the normalized PBA/AB (<2) in
MDCKII-MDR1 cells. All nine compounds examined in the trans-
port assay are highly polarized in the % cell-associated assay, but
this polarization disappears in the presence of a potent inhibitor
of Pgp.
Measurements of the % cell-associated dye (% cell) after 1 h in
the permeability assay as a fraction of compound added gave the
most intriguing data. The two weakest inhibitors of VER-induced
ATPase activity, compounds 3-S and 6-Se, showed the lowest % cell
in the absence of Pgp efflux. This may reflect the local concentra-
tion of compound in the cell in the vicinity of the Pgp ATPase stim-
ulating site, which should be reduced more as the transporter is
functioning to decrease compound concentration in the cell. How-
ever, there is no obvious relationship between % cell and ATPase
IC50 or CAM EC50. One contributing factor could be related to the
unusual polarity (>2-fold) in % cell favoring the apical surface when
the transporter is patent or uninhibited. In most cases studied for
>500 compounds of diverse structure, the % cell lacks polarity
(Sawada and Raub, unpublished observation). Upon inhibition of
Pgp, this % cell polarity disappeared and was magnitudinally sim-
ilar to the uninhibited BA % cell value. Exceptions were compounds
4-S and 4-Se that had the greatest intrinsic (Pgp inhibited) % cell at
>30%.
Compounds 7-S and 8-Te did not appear to be substrates for Pgp
based on the normalized PBA/AB of <2. However, 7-S showed the
highest passive Papp as well as the highest cell polarity in the unin-
hibited assay. In contrast, 8-Te had the lowest values of % cell in
either uninhibited or inhibited assays. In the transport assay, nei-
ther of these compounds appeared to be recognized by the pump.
It is possible that the chalcogenopyrylium compounds show both a
no-inhibitor effect32 as well as competitive inhibition of Pgp. Until
Pgp becomes saturated with either substrate or inhibitor, it is pos-
sible that neither impacts the function of the other. The range of
concentrations may vary markedly among the chalcogenopyrylium
compounds where a no-inhibitor effect is observed. As Pgp be-
comes saturated with either the substrate or inhibitor, competitive
recognition then begins. Alternatively, it may be that 8-Te is a spe-
cial case among the set, and that this molecule binds at an alloste-
ric site. This reasoning could explain why assays such as ATPase
inhibition, CAM uptake, and vinblastine transport studies display
correlative inhibition values whereas MDCKII-MDR1 studies indi-
cate that Pgp does not influence the cellular distribution of 8-Te.
In this case, competitive inhibition of ATPase relative to VER ob-
served in Figure 3 may be the result of a no-competitor effect. That
is, VER may overcome 8-Te inhibition at higher concentrations,
yet, this may not necessarily reflect overlapping binding sites.
Future studies with these systems can examine flippase versus
transporter functions and no-inhibitor versus competitive inhibi-
tion behavior over a range of concentrations and with a variety
of structural variations. The chalcogenopyrylium dyes offer a more
flexible scaffold than the chalcogenorhodamines/rosamines for
evaluating structure–activity relationships including selection
among the four chalcogen atoms (O, S, Se, and Te) in the chalco-
genopyrylium ring as well as the placement of functional groups
on alkyl and aryl substituents at the 2-, 4-, and 6-positions. Com-
pounds 1-S, 4-S, 4-Se, and 8-Te represent promising lead struc-
tures for further refinement.
Lighter chalcogen analogues of the rhodamines and pyrylium
classes, rhodamine 123 and the thiopyrylium compound AA1
(compound 6-S with E = S and R = H, Chart 2) have been evaluated
as anticancer agents.34,35 With both the chalcogen-substituted
rhodamines and the chalcogenopyrylium compounds, one might
be concerned with the toxicity associated with the heavy chalco-
gen atoms selenium and tellurium. While selenium is an essential
trace element and has been the subject of numerous toxicity stud-
ies,36,37 one might be tempted to assume that, since tellurium is
below selenium in the periodic table, tellurium must be more
metallic and consequently more toxic. Tellurium has been found
to be less toxic than selenium based on studies in the semiconduc-
tor and reprographic industries, where tellurium is in commercial
use.36,37 Organoselenium compounds, such as EbselenÒ, have
emerged as viable drug candidates38 while tellurium-containing
compounds have shown acceptable toxicity in animal studies.39
Ammonium trichloro(dioxoethylene-O,O’)tellurate (AS-101), an
inorganic tellurate complex, was shown to have both antitumor
and immunomodulatory activity in vivo.39
Onepossibleexplanationforthisobservationis that2.5 ꢀ 10ꢁ6 M
LSN335984 competes with non-specific binding of the chalcogeno-
pyrylium compounds to the apical membrane. In the presence of
inhibitor, the chalcogenopyrylium compounds are free to equili-
brate between apical and basal membranes. The polarity observed
in the inhibited system with compounds 4-S and 4-Se suggests
that these two compounds have stronger non-specific binding than
other chalcogenopyrylium compounds. However, if this is the case,
the structural similarities of 3-S, 4-S, and 4-Se as well as similar
values of logP (Table 1) suggest that non-specific binding should
be similar for all three compounds.
Another explanation is that Pgp maintains a greater concentra-
tion of the chalcogenopyrylium compounds in the apical mem-
brane, possibly in the outer leaflet, and upon inhibition of efflux
the apical and basal membranes equilibrate. This explanation is
consistent with Pgp functioning as either a flippase31 moving the
substrate from the inner to the outer leaflet or as a transporter,
where delivery to the apical surface allows diffusion back to the
outer membrane.32 In examples where Pgp is acting as a trans-
porter, high outer membrane association has been noted.32 Alter-
natively, interaction with Pgp may occur at an allosteric site that
becomes unavailable when an appropriate transport substrate (or
other inhibitor) is present. This would be a clear example of a
no-competitor effect.
Our examples do not support the idea that other transporters
are assisting Pgp in the efflux of the chalcogenopyrylium exam-
ples.33 Overall rates of diffusion in either the AB or BA directions
are slow in either the uninhibited or inhibited examples (Table
2), which argue against the active participation of additional trans-
porters. Furthermore, for 8-Te, values of EC50 for the facilitation of
CAM uptake and IC50 for [3H]-vinblastine transport in MDCKII-
MDR1 cells (5.4 ꢀ 10ꢁ6 M and 9.6 ꢀ 10ꢁ6 M, respectively) are quite
similar to values of IC50 for inhibition of VER-induced stimulation
of Pgp in lipid-activated Pgp and IC50 for inhibition of [3H]-vinblas-
4. Experimental
4.1. General methods
Solvents and reagents were used as received from Sigma–Al-
drich Chemical Co. (St. Louis, MO) unless otherwise noted. Concen-
tration in vacuo was performed on a Büchi rotary evaporator. Dry
tetrahydrofuran (THF) was obtained by distilling THF from sodium

G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
9753
benzophenone ketyl prior to use. NMR spectra were recorded on an
Inova 500 instrument with residual solvent signal as internal stan-
dard: CDCl3 (d 7.26 for proton, d 77.0 for carbon), CD2Cl2 (d 53.8 for
carbon). Infrared spectra were recorded on a Perkin-Elmer FTIR
instrument. UV–vis near-IR spectra were recorded on a Perkin-El-
mer Lambda 12 spectrophotometer or on a Shimadzu UV-3600
spectrophotometer in quartz cuvettes with a 1-cm path length.
Elemental analyses were conducted by Atlantic Microanalytical,
Inc. Compounds 1-S, 4-S and 4-Se were prepared according to
Ref. 16d. Compounds 5-S and 5-Se were prepared according to
Ref. 16b. Compounds 6-Se, 7-S, and 10-S were prepared according
to Ref. 16c. Compounds 8-S, 8-Te, 13-Te, and 14-S were prepared
according to Ref. 16a.
LSN335984 was synthesized by Eli Lilly and Co (Indianapolis,
IN). [3H]Vinblastine was obtained from Moravek Biochemicals
Inc. (Brea, CA) or GE Healthcare (Buckinghamshire, UK). Puromycin
was purchased from Sigma (St. Louis, MO) and MgCl2 from Fisher
scientific (Pittsburgh, PA). Most cell culture reagents such as
HEPES, Hank’s balanced salt solution, Dulbecco’s modified Eagle’s
medium (DMEM), DMEM/F-12, fetal bovine serum and horse ser-
um, gentamicin, phosphate-buffer saline and sucrose buffer were
ordered through Invitrogen (Carlsbad, CA). Fetal bovine serum to
coat the GF/B filter plates was purchased from Hyclone (Logan,
UT). ATP, AMP-PNP, creatine kinase, creatine phosphate and prote-
ase inhibitor cocktail tablets were obtained from Roche (Indianap-
olis, IN).
606 nm (
e
7.2 ꢀ 104 Mꢁ1 cmꢁ1). Anal. Calcd for C31H31F6N2O2PS: C,
58.12; H, 4.88; N, 4.37. Found: C, 58.31; H, 4.82; N, 4.52.
The PF6 salt (0.050 g) was dissolved in 20 mL of CH3CN and
0.250 g of Amberlite IRA-400 Chloride ion exchange resin was
added. The resulting mixture was stirred 0.5 h and was filtered to
remove the exchange resin. This process was repeated a total of
three times. After the final filtration, the filtrate was concentrated
and the residue was recrystallized from CH3CN/ether to give
0.036 g (88%) of 9-S as a metallic, purple-gold crystalline solid:
mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.52 (s, H), 8.39 (s,
1H), 8.05 (AA’BB’, 2H, J = 9 Hz), 7.92 (AA’BB’, 4H, J = 9 Hz), 7.85
(m, 2H), 7.69 (m, 3H), 7.08 (m, 4H), 3.86 (t, 8H, J = 4.6 Hz), 3.53
(t, 4H, J = 4.6 Hz), 3.46 (t, 4H, J = 4.6 Hz); kmax (CH2Cl2) 606 nm (
e
7.1 ꢀ 104 Mꢁ1 cmꢁ1). Anal. Calcd for C31H31ClN2O2S: C, 62.50; H,
5.16; N, 6.39. Found: C, 62.31; H, 5.14; N, 6.37.
4.1.3. Preparation of
thiopyran-4-one (16-S)
D-4H-2,6-bis(4-N-morpholinophenyl)-
Sulfur (0.10 g, 3.0 mmol) and NaBH4 (0.24 g, 6.2 mmol) in 60 mL
of 0.25 M NaOEt in EtOH were heated at reflux for several hours
until the solution turned clear. 1,5-Bis(4-N-morpholinophenyl)-
1,4-pentadiyn-3-one (1.00 g, 2.5 mmol)16c was stirred in 0.25 M
NaOEt in EtOH (60 mL) at 50 °C until completely dissolved. The
resulting solution was then added to the solution of Na2S at reflux.
The reaction mixture was heated at reflux for 2 h and was then
concentrated. Water (150 mL) was added and the products were
extracted with CH2Cl2 (3 ꢀ 50 mL). The combined organic extracts
were dried over Na2SO4 and concentrated. The crude product was
recrystallized from EtOAc to give 0.87 g (80%) of thiapyranone 16-
S, mp 218–221 °C; 1H NMR [500 MHz, CDCl3] d 7.56 (AA’BB’, 4H, J =
9 Hz), 7.13 (s, 2H), 6.94 (AA’BB’, 4H, J = 9 Hz), 3.86 (t, 8H, J = 4.8 Hz),
3.25 (t, 8H, J = 4.8 Hz); 13C NMR [126 MHz, CDCl3] d 182.8, 152.7,
152.6, 127.8, 126.5, 124.8, 114.9, 66.6, 48.0; IR (KBr) 2960, 2835,
1509 cmꢁ1; HRMS (EI) m/z 434.1678 (Calcd for C25H26N2O3S:
434.1678).
4.1.1. Preparation of
D-4H-2-(4-N-morpholinophenyl)-6-phenyl-
thiopyran-4-one (15-S)
Sulfur (0.12 g, 3.8 mmol) and NaBH4 (0.30 g, 7.9 mmol) in 60 mL
of 0.25 M NaOEt in EtOH were heated at reflux for several hours
until the solution turned clear. 1-(4-N-Morpholino-phenyl)-5-phe-
nyl-1,4-pentadiyn-3-one (1.00 g, 3.2 mmol)16c was stirred in
0.25 M NaOEt in EtOH (60 mL) at 50 °C until completely dissolved.
The resulting solution was then added to the solution of Na2S at re-
flux. The reaction mixture was heated at reflux for 2 h and was
then concentrated. Water (150 mL) was added and the products
were extracted with CH2Cl2 (3 ꢀ 50 mL). The combined organic ex-
tracts were dried over Na2SO4 and concentrated. The crude product
was purified via chromatography on SiO2 (20% EtOAc/CH2Cl2) and
was recrystallized from EtOAc to give 0.43 g (33%) of thiapyranone
4.1.4. Preparation of 2,6-bis(4-N-morpholinophenyl)-4-phenyl-
thiopyrylium chloride (11-S)
Phenylmagnesium bromide (0.58 mL of
a 1.0 M solution,
0.58 mmol) was added dropwise to a solution of thiopyranone
16-S (0.050 g, 0.12 mmol) in THF (2.5 mL) and the resulting solu-
tion was heated at reflux for 2 h. The reaction mixture was cooled
and poured into acetic acid (3 mL). Hexafluorophosphoric acid
(60% by weight in water) was added dropwise until a color change
was observed. Water (10 mL) was added and the resulting solution
was cooled to ꢁ10 °C. The resulting precipitate was collected by fil-
tration and the solid was washed with water (10 mL) and ether
(10 mL). The crude product was recrystallized from CH3CN and
ether to give 0.070 g (91%) of 11-S (PF6ꢁ) as a dark blue solid,
mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.30 (s, 2H), 7.90 (m,
6H), 7.70 (m, 3H), 7.06 (AA’BB’, 4H, J = 9 Hz), 3.86 (t, 8H, J =
I
15-S, mp 154–156 °C; H NMR [500 MHz, CDCl3] d 7.60 (m, 7H),
7.08 (s, 2H), 6.93 (AA’BB’, 2H, J = 9 Hz), 3.79 (t, 4H, J = 4.8 Hz),
3.20 (t, 4H, J = 4.8 Hz); 13C NMR [126 MHz, CDCl3] d 182.3, 153.3,
152.9, 152.5, 136.6, 130.8, 129.5, 127.9, 127.2, 127.1, 126.3,
124.9, 115.1, 66.9, 48.3; IR (KBr) 3126, 2962, 1610 cm-1; HRMS
(EI) m/z 349.1136 (Calcd for C21H19NO2S: 349.1120).
4.1.2. Preparation of 2,4-bis(4-N-morpholinophenyl)-6-phenyl-
thiopyrylium chloride (9-S)
A solution of
ran-4-one 15-S (0.10 g, 0.29 mmol) in THF (5 mL) was added drop-
wise to solution of 4-bromo-N-phenylmorpholine (0.35 g,
D-4H-2(4-N-morpholinophenyl)-6-phenylthiopy-
4.8 Hz), 3.48 (t, 8H, J = 4.8 Hz); kmax (CH2Cl2) 643 nm (
e
6.6 ꢀ 104
a
Mꢁ1 cmꢁ1). Anal. Calcd for C31H31F6N2O2PS: C, 58.12; H, 4.88; N,
4.37. Found: C, 58.00; H, 4.83; N, 4.41.
1.4 mmol) and magnesium turnings (0.070 g, 2.8 mmol) in 1 mL
of dry THF. The resulting mixture was heated at reflux for 2 h,
cooled, and then poured into acetic acid (3 mL). Hexafluorophos-
phoric acid (60% by weight in water) was added dropwise until a
color change was observed. Water (10 mL) was added and the
resulting solution was cooled to ꢁ10 °C. The resulting precipitate
was collected by filtration and the solid was washed with water
(10 mL) and ether (10 mL). The crude product was recrystallized
from CH3CN and ether to give 0.16 g (86%) of 9-S (PF6ꢁ) as a dark
blue solid, mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.52 (s, H),
8.39 (s, 1H), 8.05 (AA’BB’, 2H, J = 9 Hz), 7.92 (AA’BB’, 4H, J =
9 Hz), 7.85 (m, 2H), 7.69 (m, 3H), 7.08 (m, 4H), 3.86 (t, 8H, J =
4.6 Hz), 3.53 (t, 4H, J = 4.6 Hz), 3.46 (t, 4H, J = 4.6 Hz); kmax (CH2Cl2)
The PF6 salt (0.050 g) was dissolved in 20 mL of CH3CN and
0.250 g of Amberlite IRA-400 Chloride ion exchange resin was
added. The resulting mixture was stirred 0.5 h and was filtered to
remove the exchange resin. This process was repeated a total of
three times. After the final filtration, the filtrate was concentrated
and the residue was recrystallized from CH3CN/ether to give
0.038 g (92%) of 11-S as a metallic, purple-gold crystalline solid:
mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.30 (s, 2H), 7.90 (m,
6H), 7.70 (m, 3H), 7.06 (AA’BB’, 4H, J = 9 Hz), 3.86 (t, 8H, J =
4.8 Hz), 3.48 (t, 8H,
J = 4.8 Hz); kmax (CH2Cl2) 643 nm (e
6.6 ꢀ 104 Mꢁ1 cmꢁ1). Anal. Calcd for C31H31ClN2O2S: C, 70.10; H,
5.88; N, 5.27. Found: C, 70.03; H, 5.81; N, 5.31.

9754
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
4.1.5. Preparation of 2,6-bis(4-N-morpholinophenyl)-4-(4-N,N-
identical except that each reaction contained 2 lg of lipid-acti-
dimethylaminophenyl)-selenopyrylium chloride (12-Se)
vated Pgp and IC50 values were determined in the presence of
A solution of
D
-4H-2,6-bis(4-N-morpholinophenyl)-selenopy-
2 ꢀ 10ꢁ4 M VER.
ran-4-one 16-Se (0.10 g, 0.21 mmol)16c in THF (5 mL) was added
dropwise to a solution of 4-bromo-N,N-dimethylaniline (0.20 g,
1.0 mmol) and magnesium turnings (0.040 g, 1.7 mmol) in 1 mL
of dry THF. The resulting mixture was heated at reflux for 2 h,
cooled, and then poured into acetic acid (3 mL). Hexafluorophos-
phoric acid (60% by weight in water) was added dropwise until a
color change was observed. Water (10 mL) was added and the
resulting solution was cooled to ꢁ10 °C. The resulting precipitate
was collected by filtration and the solid was washed with water
(10 mL) and ether (10 mL). The crude product was recrystallized
from CH3CN and ether to give 0.15 g (97%) of 12-Se (PF6ꢁ) as a blue
solid, mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.29 (s, 2H), 8.03
(AA’BB’, 2H, J = 9 Hz), 7.81 (AA’BB’, 4H, J = 9 Hz), 7.07 (AA’BB’, 4H,
J = 9 Hz), 6.94 (AA’BB’, 2H, J = 9 Hz), 3.89 (t, 8H, J = 4 Hz), 3.44 (t,
4.4. Fluorescence assays
Pgp (cys-less) was lipid-activated as described for ATPase stud-
ies.13 Lipids were added at a ratio of 2:1 (w/w) vs. Pgp at a final
concentration of 0.2%. Dilutions of the chalcogenopyrylium dyes
were added to the microplate (1 lL each) in DMSO (final concen-
tration of DMSO was 2%). Activated Pgp was added to a final con-
centration of 2.5 ꢀ 10ꢁ7 M in TE buffer (0.040 M Tris–HCl pH 7.4,
1 ꢀ 10ꢁ4 M EGTA) to a final volume of 50
lL. Reactions were set
up in the dark (minimum indirect light) and the microplate was
wrapped in aluminum foil and incubated at 37 °C for 10 min. Con-
trol reactions that contained the identical concentration of E. coli
lipids, but lacking Pgp, were set up in parallel. Fluorescence was
measured from the top in a Spectramax M5 Plate Reader at
ambient temperature with kEX = 584 nm and kEM = 690 nm with a
630-nm cutoff filter. Data (rfu) were plotted using SigmaPlot
2000 software and the fluorescence increase in the presence of
8H, J = 4 Hz), 3.22 (s, 6H); kmax (CH2Cl2) 636 nm (
e
4.7 ꢀ 104
M
ꢁ1 cmꢁ1). Anal. Calcd for C33H36F6N3O2PSe: C, 52.95; H, 5.12; N,
5.61. Found: C, 53.25; H, 5.15; N, 5.50.
The PF6 salt (0.050 g) was dissolved in 20 mL of CH3CN and
0.250 g of Amberlite IRA-400 Chloride ion exchange resin was
added. The resulting mixture was stirred 0.5 h and was filtered to
remove the exchange resin. This process was repeated a total of
three times. After the final filtration, the filtrate was concentrated
and the residue was recrystallized from CH3CN/ether to give
0.038 g (91%) of 11-S as a metallic, purple-gold crystalline solid:
mp >260 °C; 1H NMR [500 MHz, CD2Cl2] d 8.29 (s, 2H), 8.03 (AA’BB’,
2H, J = 9 Hz), 7.81 (AA’BB’, 4H, J = 9 Hz), 7.07 (AA’BB’, 4H, J = 9 Hz),
6.94 (AA’BB’, 2H, J = 9 Hz), 3.89 (t, 8H, J = 4 Hz), 3.44 (t, 8H, J = 4 Hz),
Pgp was fit to a hyperbolic curve to determine Kapp
.
4.5. Enhancement of calcein-AM uptake into MDCKII-MDR1
cells
MDCK cells transfected with wild-type MDR1 (ABCB1)23 were
obtained at passage number 12 from Dr. Piet Borst at The
Netherlands Cancer Institute. Cell growth was maintained in Dul-
becco’s Modified Eagle’s Medium (Gibco) supplemented with 10%
3.22 (s, 6H); kmax (CH2Cl2) 636 nm (
Calcd for C33H36ClN3O2Se: C, 63.82; H, 5.84; N, 6.77. Found: C,
e
4.6 ꢀ 104 Mꢁ1 cmꢁ1). Anal.
fetal bovine serum (FBS), 100 U/mL penicillin, and 100 lg/mL
streptomycin in 75-cm2 flasks. Cultures were passaged by trypsin-
ization 1:10 twice a week and used at Passage Nos. 16–42. Cells
were seeded at 40,000 cells/well in 96-well flat bottom plates
63.65; H, 6.01; N, 6.78.
4.2. Determination of partition coefficients
(Falcon) using a medium volume of 200
on day 3 prior to their use on day 4.
lL, which was replaced
The octanol/water partition coefficients were measured using
UV–vis spectrophotometry and the ‘shake flask’ direct measure-
ment.40 Mixing for 1 h was followed by 4 h of settling time.
Equilibration and measurements were made at 23 °C using a Per-
kin-Elmer Lambda 12 spectrophotometer. HPLC-grade 1-octanol
(Sigma–Aldrich) and deionized water were used.
For the two-dose inhibition screen, cells were washed once
with Dulbecco’s phosphate-buffered saline containing 0.010 M
Hepes buffer at pH 7.4 (DPBSH) (Gibco) and incubated in the
dark with 80
l
L of 5 ꢀ 10ꢁ6 M or 2.5 ꢀ 10ꢁ5 M chalcogenopyryli-
um compound in DPBSH at 37 °C. For the calculation of values of
EC50, cells were washed once with DPBSH and incubated with
various concentrations of chalcogenopyrylium compound in
DPBSH at 37 °C in the dark. Concentrations of the various com-
pounds came from 1:1 serial dilution series with a maximum
concentration of 1 ꢀ 10ꢁ4 M. After 30 min, the test compound
4.3. ATPase assays
ATP hydrolysis determined by the spectrophotometric-coupled
assay was performed in microplate format in 96-well plates as pre-
was replaced to include 0.5 l
g mLꢁ1 CAM and incubated an
viously described.13 Each reaction contained 4–5
l
g of lipid-acti-
vated Cys-less mouse mdr3 Pgp with the indicated amount of
drug or chalcogenopyrylium compound added in 1- L volume
additional 20 min in the dark. Calcein fluorescence was read on
a Cytofluor series 4000 Multi Well Plate Reader (PerSeptive
Biosystems) with kEX and kEM set at 485 nm and 530 nm,
respectively. Negative (0.25% DMSO in DPBSH), and positive
(2.5 ꢀ 10ꢁ6 M LSN335984, a specific Pgp inhibitor) controls were
included in each plate. EC50 values were calculated from the se-
rial dilution curves using GraphPad PRISM version 4.03 software.
l
from concentrated DMSO stock solutions. Cys-less protein was
routinely used for ATPase measurements for consistency and to
avoid DTT, which is necessary to activate wild-type Pgp, but was
found to reduce some of the compounds. DMSO was 2% final in all
reactions. Each 50-
l
L reaction contained 0.040 M Tris–HCl pH 7.4,
Briefly, compound concentration was plotted as log lM concen-
1 ꢀ 10ꢁ4 M EGTA, 0.010 M NaATP, 0.012 M MgSO4, 1 ꢀ 10ꢁ3
M
tration versus relative fluorescence units (rfu) and a sigmoidal
dose-response (variable slope) analysis with no weighting or
restrictions was applied.
PEP, 1.5 ꢀ 10ꢁ3 M NADH, and pyruvate kinase and lactate dehydro-
genase (each at a final concentration of 0.1 mg/mL). Reactions were
kept on ice and in the dark until placed in the plate reader. Kinetics
of NADH oxidation was followed at 37 °C by a decrease in absorp-
tion at 340 nm and this was converted into a specific activity for
moles ATP hydrolyzed. Control reactions containing DMSO alone
or 1.5 ꢀ 10ꢁ4 M VER were performed in parallel for comparison.
Assays were performed in a Spectramax Gemini plate reader and
analyzed with SOFTmaxPro software. Studies of the inhibition of
VER-stimulated ATPase (for IC50 values) activity were essentially
4.6. Pgp-transport studies across MDCKII-MDR1 monolayers
MDCKII-MDR1 cells that were seeded at 5 ꢀ 104 cells cm-2 onto
12-well (1.13 cm2 surface area) Transwell polycarbonate filters
(Costar) were fed on days 3 and 5, and used on day 6. The upper
and lower chamber volumes were 0.5 mL and 1.0 mL, respectively.
Cells were rinsed 10 min in DPBSH at 37 °C with mixing on a nuta-

G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
9755
tor (Clay Adams). Cells were pre-incubated with 4.3 mg mLꢁ1
4.9. Isolation of membrane vesicles
bovine serum albumin (BSA) in DPBSH alone or containing
2.5 ꢀ 10ꢁ6 M LSN335984. After 30 min, 5 ꢀ 10ꢁ6 M test compound
in BSA/DPBSH with or without inhibitor was added to the donor
chamber (0.5 mL upper or apical, 1.0 mL lower or basolateral). Ini-
tial donor samples were taken at t = 0. For apical-to-basolateral (A–
B) flux, D0 was taken from the mixing tube before addition to the
cell monolayer. For basolateral-to-apical (B–A) flux this sample
was taken from the 12-well plate 10 min after transfer, but before
cell wells were added. Samples were taken from both the donor
and receiver chambers following a 1-h incubation in the dark at
37 °C with constant mixing by nutation. Cell monolayers were
rinsed briefly two times using cold DPBS and extracted with
Plasma membranes were prepared by nitrogen decompression
and differential centrifugation as described by Lever26 with modi-
fication. Cultured PEAK-MDR1 cells were centrifuged at 800g for
5 min at 4 °C. Cells were re-suspended in 50 mL phosphate-buffer
saline, pH 7.4 at 4 °C, and washed once with 50 mL 2 ꢀ 10ꢁ4 M
MgCl2-sucrose buffer by centrifugation at 800g for 5 min at 4 °C.
The cell pellet was washed once with 2 ꢀ 10ꢁ4 M CaCl2-sucrose
buffer supplemented with protease inhibitor cocktail (1 protease
inhibitor cocktail tablet per 50 ml CaCl2-sucrose buffer) before
equilibration for 30 min at 150 p.s.i. of nitrogen pressure in the cell
disruption bomb (Parr Instrument Co., Moline, IL). Cells were burst
by slow release of pressure. After removal of nuclei and unbroken
cells by centrifugation at 800g for 5 min at 4 °C, the supernatant
was supplemented with 0.001 M EDTA. The supernatant was lay-
ered onto a 35% sucrose buffer cushion containing 0.01 M Tris,
0.001 M EDTA, and pH 7.4 and centrifuged at 16,000g for 1 h at
4 °C. Membranes collected at the interface were subsequently pel-
leted at 100,000g for 1 h at 4 °C, re-suspended in 0.25 M sucrose,
and passed through a 25-ga. needle to vesicularize the membrane
preparation.27 The prepared membrane vesicles were stored in a li-
quid nitrogen freezer. The protein concentration of the vesicle
preparation was determined using the bicinchoninic acid method
described by Smith et al.42
500 lL methanol for 3 min.
The concentration of analytes was determined by LC-MS/MS
after dilution of samples with equal part methanol. Chromatog-
raphy was effected on a Javelin BETASIL C18 column (20 ꢀ 2.1
mm, 5 l
) at a flow rate of 1.5 mL minꢁ1 using a gradient of
40% mobile phase B with a hold time of 0.1 min, followed by a
0.1 min linear gradient to 80% B with a hold time of 0.15 min,
followed by a step gradient to 98% B with a hold time of
0.4 min where mobile phase
A is water/TFA/1 M NH4HCO3
(20,000:8:2, v/v) and mobile phase B is ACN/TFA/1 M NH4HCO3
(2000:8:2, v/v) Column eluent was diverted to waste from 0 to
0.15 min and then introduced into a Sciex API 4000 triple quad-
rupole mass spectrometer with turboionspray source operated in
the positive mode at 740 °C for 0.3 min. MRM experiments mon-
itored were: 349/333 for 1-S; m/z 377/361 for 3-S; m/z 391/375
for 4-S; m/z 439/423 for 4-Se; m/z 383/367 for 5-S; m/z 431/351
for 5-Se; m/z 446/366 for 6-Se; m/z 427/351 for 7-S; and m/z
551/134 for 8-Te.
The inhibition assay was conducted using 1 ꢀ 10ꢁ7 M [3H]-vin-
blastine (0.25 l
Ci mLꢁ1) as a Pgp substrate. The accumulation of
[3H]-vinblastine into Pgp membrane vesicles was inhibited by
8-Te (2.5 ꢀ 10ꢁ7 M–2.5 ꢀ 10ꢁ5 M), using 5 ꢀ 10ꢁ6 M LSN335984
as a standard for complete Pgp inhibition. Previously, the inhibition
of [3H]-vinblastine accumulation in Pgp membrane vesicles by
LSN335984 was shown to exhibit an IC50 value of (6.5 1.1) ꢀ
10ꢁ9 M and thus a complete inhibition of Pgp was assumed at
5 ꢀ 10ꢁ6 M. The reaction mixtures contained 0.004 M ATP, ATP
with different concentrations of 8-Te (2.5 ꢀ 10ꢁ7 M–2.5 ꢀ
10ꢁ5 M), ATP with 5 ꢀ 10ꢁ6 M LSN335984, or 0.004 M AMP-PNP.
All reaction mixtures contained 1 ꢀ 10ꢁ7 M [3H]-vinblastine, the
4.7. Inhibition of vinblastine efflux in MDCKII-MDR1 cells by
8-Te
MDCKII-MDR1 cells were seeded onto Costar Transwell polycar-
bonate membranes and maintained as previously described. On
day 6, cells were rinsed 1 ꢀ 10 min in DPBSH with nutation at
37 °C, and then conditioned in 0, 0.1, 0.5, 1, 5, 10, 20, or 40 ꢀ 10
ꢁ6 M 8-Te in BSA/DPBSH in the dark. After 30 min at 37 °C,
regenerating system including 100
0.010 M creatine phosphate, and 0.010 M MgCl2. The reaction
was started by addition of 30 L reaction mixtures to vesicles
(20 g/20 L) or 20 L sucrose buffer. The reactions were incu-
l
g mLꢁ1 creatine kinase,
l
1.0 mL [3H]-vinblastine (0.25
l
Ci mLꢁ1 from 0.1 mCi mLꢁ1 EtOH
l
l
l
stock) in appropriate 8-Te solution was introduced to the basolat-
eral chamber and 0.5 mL fresh 8-Te was added to the apical cham-
ber. The incubator was maintained in the dark during experiments.
Initial donor samples were taken from the basolateral chamber at
t = 0. The apical solution was replaced every 10 min with fresh buf-
fer and appearance of [3H]-vinblastine in the apical chamber was
measured by scintillation counting. An EC50 was calculated using
XLfit software.
bated for 2.5 min at 37 °C with gentle agitation in the dark. The
reactions were stopped by the addition of ice-cold sucrose buffer
to wells and filtered onto a membrane filter (GF/B glass fiber plate
soaked overnight in 10% FBS/0.25 M sucrose buffer, v/v) with a 96-
channel cell harvester (PerkinElmer, Boston, MA). Radioactivity re-
tained on the filter was determined by liquid scintillation counting.
4.10. Calculations for inhibition studies
4.8. Inhibition of vinblastine efflux in a membrane vesicle model
by 8-Te
All values were corrected for background. The background was
determined by the accumulation of [3H]-vinblastine in the pres-
ence of 5 ꢀ 10ꢁ6 M LSN335984, which completely inhibits [3H]-
vinblastine uptake mediated by Pgp. The calculation of the concen-
tration of inhibitor resulting in 50% inhibition (IC50 value) was then
determined by nonlinear regression analysis using WinNonlin Pro-
Human embryonic kidney (PEAKSTABLE) cells over expressing
human Pgp were obtained from William Perry at Eli Lilly and
Company. The detailed transfection procedure for development
of this cell line was described by Godinot et al.41 PEAK-MDR1
cells were grown in 5% CO2 at 37 °C in a humidified atmosphere
fessional, version 5.0.1 (Pharsight Corporation, Mountain View,
DPMmax
CA), where
%
activity in relation to control
=
100 ꢀ ð1þ
,
½Iꢃ
Þs
IC
in DMEM supplemented with 10% fetal bovine serum, 50
lg/mL
50
[I] = inhibitor concentration, DPMmax is the highest [3H]-vinblas-
gentamicin, and 0.5 g/mL puromycin. Cells were grown in mon-
l
tine uptake with background corrected, and s is the slope factor.
olayers until they reached 80% confluency and passaged twice
before being adopted to suspension culture. Cells grown in sus-
pension were cultured in spinner flasks containing modified
DMEM/F-12 medium supplemented with 2% horse serum and
Acknowledgments
This research was supported in part by NIH Grant T32 CA09363
(Postdoctoral Training Grant) to G.T. G.T. gratefully acknowledges
0.3
lg/mL puromycin until total cell number was approximately
109 cells.

9756
G. A. Sawada et al. / Bioorg. Med. Chem. 16 (2008) 9745–9756
18. Tombline, G.; Holt, J. J.; Gannon, M. K.; Donnelly, D. J.; Wetzel, B.; Sawada, G. A.;
Raub, T. J.; Detty, M. R. Biochemistry 2008, 47, 3294.
19. (a) Al-Shawi, M. K.; Polar, M. K.; Omote, H.; Figler, R. A. J. Biol. Chem. 2003, 278,
52629; (b) Omote, H.; Fiegler, R. A.; Polar, M. K.; Al-Shawi, M. K. Biochemistry
2004, 43, 3917.
20. Ohulchanskyy, T. Y.; Gannon, M. K., II; Ye, M.; Wagner, S. J.; Skripchenko, A.;
Prasad, P. N.; Detty, M. R. J. Phys. Chem. B 2007, 111, 9686.
21. (a) Qu, Q.; Chu, J. W.; Sharom, F. J. Biochemistry 2000, 42, 1345–1353; (b) Lugo,
M. R.; Sharom, F. J. Biochemistry 2005, 44, 14020–14029; (c) Lugo, M. R.;
Sharom, F. J. Biochemistry 2005, 44, 643–655.
Alan Senior, Robert Bambara, and Barbara Iglewski for encourage-
ment and the use of their laboratory space and supplies. We thank
Dr. Piet Borst at The Netherlands Cancer Institute for supplying the
MDCKII-MDR1 cells and William Perry at Eli Lilly and Company for
the human embryonic kidney (PEAKSTABLE) cells.
References and notes
22. Litman, T.; Zeuthen, T.; Skovsgaard, T.; Stein, W. D. Biochim. Biophys. Acta 1997,
1361, 169.
23. Evers, R.; Kool, M.; Smith, A. J.; van Deemter, L.; de Haas, M.; Borst, P. Br. J.
Cancer 2000, 83, 366.
1. (a) Gottesman, M. M.; Fojo, T.; Bates, S. E. Nature Rev./Cancer 2002, 2, 48; (b)
Szakacs, G.; Paterson, J. K.; Ludwig, J. A.; Booth-Genthe, C.; Gottesman, M. M.
Nature Rev./Drug Discov. 2006, 5, 219.
2. Dean, M.; Rzhetsky, A.; Allikmets, R. Genome Res. 2001, 11, 1156.
3. (a) Seelig, A.; Gatlik-Landwojtowicz, E. Mini-Rev. Med. Chem. 2005, 5, 135; (b)
Raub, T. J. Mol. Pharm. 2006, 3, 3.
4. Burger, H.; Foekens, J. A.; Look, M. P.; Meijer-van Gelder, M. E.; Klijn, J. G. M.;
Wiemer, E. A. C.; Stoter, G.; Nooter, K. Clin. Cancer Res. 2003, 9, 827.
5. Loscher, W.; Potschka, H. Nat. Rev. Neurosci. 2005, 6, 591.
6. (a) Tiberghien, F.; Loor, F. Anti-Cancer Drugs 1996, 7, 568; (b) Hirsch-Ernst, K. I.;
Ziemann, C.; Rustenbeck, I.; Kahl, G. F. Toxicology 2001, 167, 47; (c) Loor, F.;
Tiberghien, F.; Wenandy, T.; Didier, A.; Traber, R. J. Med. Chem. 2002, 45, 4598.
7. Fisher, G. A.; Lum, B. L.; Hausdorff, J.; Sikic, B. I. Euro. J. Cancer 1996, 32A, 1082.
8. Fojo, T.; Bates, S. Oncogene 2003, 22, 7512.
9. (a) Lee, J. S.; Paull, K.; Alvarez, M.; Hose, C.; Monks, A.; Grever, M.; Fojo, A. T.;
Bates, S. E. Mol. Pharm. 1994, 46, 627; (b) Scala, S.; Akhmed, N.; Rao, U. S.; Paull,
K.; Lan, L.-B.; Dickstein, B.; Lee, J.-S.; Elgemeie, G. H.; Stein, W. D.; Bates, S. E.
Mol. Pharm. 1997, 51, 1024.
10. (a) Eytan, G. D.; Regev, R.; Hurwitz, C. D.; Assaraf, Y. G. Eur. J. Biochem. 1997, ,
248, 104; (b) Lu, P.; Liu, R.; Sharom, F. J. Eur. J. Biochem. 2001, 268, 1687; (c)
Loetchutinat, C.; Saengkhae, C.; Marbeuf-Gueye, C.; Garnier-Suillerot, A. Eur. J.
Biochem. 2003, 270, 476.
24. Troutman, M. D.; Thakker, D. R. Pharm. Res. 2003, 20, 1192–1199.
25. Sawada, G. A.; Barsuhn, C. L.; Lutzke, B. S.; Houghton, M. E.; Padbury, G. E.; Ho,
N. F. H.; Raub, T. J. J. Pharm. Exp. Therap. 1999, 288, 1317–1326.
26. Lever, J. E. Biochem. Biophys. Res. Commun. 1977, 79, 1051.
27. Dantzig, . A. H.; Shepard, R. L.; Cao, J.; Law, K. L.; Ehlhardt, W. J.; Baughman, T.
M.; Bumol, T. F.; Starling, J. J. Cancer Res. 1996, 56, 4171.
28. Shapiro, A. B.; Ling, V. Eur. J. Biochem. 1997, 250, 130–137.
29. Loo, T. W.; Bartlett, M. C.; Clarke, D. M. J. Biol. Chem. 2003, 278, 50136.
30. Loo, T. W.; Bartlett, M. C.; Clarke, D. M. J. Biol. Chem. 2007, 22, 32043.
31. (a) Eckford, P. D. W.; Sharom, F. J. Biochem. J. 2005, 389, 517; (b) Sharom, F. J.
Biochem. Cell Biol. 2006, 84, 979.
32. Acharya, P.; Tran, T. T.; Polli, J. W.; Ayrton, A.; Ellens, H.; Bentz, J. Biochemistry
2006, 45, 15505.
33. Acharya, P.; O’Connor, M. P.; Polli, J. W.; Ayrton, A.; Ellens, H.; Bentz, J. P. Drug
Metab. Dispos. 2008, 36, 452.
34. Bernal, S. D.; Lampdis, T. J.; McIsaac, R. M.; Chen, L. B. Science 1986, 222, 169.
35. Sun, X.; Wong, J. R.; Song, K.; Hu, J.; Garlid, K. D.; Chen, L. B. Cancer Res. 1994,
54, 1465.
36. Fishbein, L. In Metals and Their Compounds in the Environment; Merian, E., Ed.;
VCH Publishers: New York, 1991; pp 1211–1226. Chapter II.28.
37. Longnecker, M. P.; Taylor, P. R.; Levander, O. A.; Howe, M.; Veillon, C.; McAdam,
P. A.; Patterson, K. Y.; Holden, J. M.; Stampfer, M. J.; Morris, J. S. Am. J. Clin. Nutr.
1991, 53, 1288.
11. Pajeva, I. K.; Wiese, M. J. Med. Chem. 2002, 45, 5671.
12. (a) Raub, T. J. Mol. Pharm. 2006, 3, 3; (b) Li, W.-X.; Li, L.; Eksterowicz, J.; Ling, X.
B.; Cardozo, M. J. Chem. Inf. Model. 2007, 47, 2429.
13. Tombline, G.; Donnelly, D. J.; Holt, J. J.; You, Y.; Ye, M.; Gannon, M. K.; Nygren,
C. L.; Detty, M. R. Biochemistry 2006, 45, 8034.
38. Mugesh, G.; DuMont, W.-W.; Sies, H. Chem. Rev. 2001, 101, 2125.
39. (a) Nyska, A.; Waner, T.; Pirak, M.; Albeck, M.; Sredni, B. Arch. Toxicol. 1989, 63,
386–393; (b) Sredni, B.; Caspi, R. R.; Klein, A.; Kalechman, Y.; Danziger, Y.;
BenYa’akov, M.; Tamari, T.; Shalit, F.; Albeck, M. Nature 1987, 330, 173–176; c
Albeck, M.; Pavliv, L.; Sredni, B. New Complexes of Tellurium and Selenium
Compounds Having Improved Solubility, for Induction of Cytokine(s). DK
Patent 167148B, 1993.
40. Sangster, J. In Octanol–Water Partition Co-efficients: Fundamentals and Physical
Chemistry; Fogg, P. G. T., Ed.; John Wiley and Sons: New York, 1997.
41. Godinot, N.; Iversen, P. W.; Tabas, L.; Xia, X.; Williams, D. C.; Dantzig, A. H.;
Perry, W. L. Mol. Cancer Ther. 2003, 2, 307.
14. Holt, J. J.; Gannon, M. K.; Tombline, G.; McCarty, T. A.; Page, P. M.; Bright, F. V.;
Detty, M. R. Biorg. Med. Chem. 2006, 14, 8635.
15. Powers, S. K.; Walstad, D. L.; Brown, J. T.; Detty, M. R.; Watkins, P. J. J. Neuro-
Oncol. 1989, 7, 179–185.
16. (a) Leonard, K. A.; Nelen, M. I.; Simard, T. P.; Davies, S. R.; Gollnick, S. O.; Oseroff, A.
R.; Gibson, S. L.; Hilf, R.; Chen, L. B.; Detty, M. R. J. Med. Chem. 1999, 42, 3953–
3964; (b) Leonard, K. A.; Hall, J. P.; Nelen, M. I.; Davies, S. R.; Gollnick, S. O.;
Oseroff, A. R.; Gibson, S. L.; Hilf, R.; Chen, L. B.; Detty, M. R. J. Med. Chem. 2000, 43,
4488–4498; (c) Brennan, N. K.; Hall, J. P.; Davies, S. R.; Gollnick, S. O.; Oseroff, A.
R.; Gibson, S. L.; Hilf, R.; Detty, M. R. J. Med. Chem. 2002, 45, 5123–5135; (d)
McKnight, R. E.; Ye, M.; Ohulchanskyy, T. Y.; Sahabi, S.; Wetzel, B. R.; Wagner, S. J.;
Skripchenko, A.; Detty, M. R. Bioorg. Med. Chem. 2007, 15, 4406–4418.
17. Tombline, G. L.; Urbatsch, I. L.; Virk, N.; Muharemagic, A.; Bartholomew White,
L.; Senior, A. E. Arch. Biochem. Biophys. 2006, 445, 124.
42. Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.;
Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal.
Biochem. 1985, 150, 76.