11C- and 18F-Labeled Radioligands for P-Glycoprotein Imaging by Positron Emission Tomography

Article abstract of DOI:10.1002/cmdc.201500420

P-Glycoprotein (P-gp) is an efflux transporter widely expressed at the human blood-brain barrier. It is involved in xenobiotics efflux and in onset and progression of neurodegenerative disorders. For these reasons, there is great interest in the assessmen

Full text of DOI:10.1002/cmdc.201500420

DOI: 10.1002/cmdc.201500420
Full Papers
1
1
18
C- and F-Labeled Radioligands for P-Glycoprotein
Imaging by Positron Emission Tomography
[a, b]
[a]
[a]
[a]
Mariangela Cantore,*
Rudi A. J. O. Dierckx, Nicola Antonio Colabufo,
Marcel Benadiba, Philip H. Elsinga, Chantal Kwizera,
[a]
[b, c]
[a]
and Gert Luurtsema
P-Glycoprotein (P-gp) is an efflux transporter widely expressed
at the human blood–brain barrier. It is involved in xenobiotics
efflux and in onset and progression of neurodegenerative dis-
orders. For these reasons, there is great interest in the assess-
ment of P-gp expression and function by noninvasive tech-
niques such as positron emission tomography (PET). Three ra-
thoxyphenyl)oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetrahy-
18
droisoquinoline ([ F]-7), were tested in several in vitro biologi-
cal assays to assess the effect of the aryl substituent in terms
of potency and mechanism of action toward P-gp. Methyl de-
11
rivative [ C]-5 is a potent P-gp substrate, whereas the corre-
18
sponding fluoroethyl derivative [ F]-7 is a P-gp inhibitor. Fluo-
11
18
diolabeled aryloxazole derivatives: 2-[2-(2-methyl-( C)-5-me-
thoxyphenyl)oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetrahy-
romethyl compound [ F]-6 is classified as a non-transported P-
gp substrate, because its efflux increases after cyclosporine A
modulation. These studies revealed a promising substrate and
11
18
droisoquinoline
([ C]-5);
2-[2-(2-fluoromethyl-( F)-5-
11
18
methoxyphenyl)oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetra-
inhibitor, [ C]-5 and [ F]-7, respectively, for in vivo imaging of
18
18
hydroisoquinoline ([ F]-6); and 2-[2-(2-fluoroethyl-( F)-5-me-
P-gp by using PET.
Introduction
[
6]
P-Glycoprotein (P-gp, ABCB1) is the best-known and most-
loid plaques, the major hallmark of Alzheimer’s disease. For
these reasons, there is great interest in the assessment of P-gp
expression and function by noninvasive techniques such as
positron emission tomography (PET).
[1]
studied ATP binding cassette (ABC) transporter. This protein
is widely expressed in barriers and excretory tissues, in which it
modulates the entry of a broad range of structurally unrelated
compounds by active ATP-driven efflux transport. P-gp is
highly expressed on the luminal membrane of the endothelial
11
11
Radiolabeled P-gp substrates such as (R)-[ C]verapamil ([ C]-
[
7,8]
11
[9]
1)
and [ C]-N-desmethylloperamide have been used to
[2]
cells at the blood–brain barrier. Owing to this crucial localiza-
tion, it plays an important role in maintaining homeostasis in
the central nervous system by protecting the brain from accu-
mulation of toxic compounds by extruding them from the
brain into the blood.
assess P-gp functional activity. Several studies have been per-
11
[9]
formed with [ C]-1 (Figure 1), which is the only tracer so far
used in clinical studies. Nevertheless, the low signal-to-noise
ratio (i.e., modest increase in brain uptake after P-gp inhibi-
[
10,11]
tion)
limits its use.
Recently, scientific evidence demonstrated that changes in
P-gp function and/or expression are involved in the onset and
progression of neurodegenerative diseases such as Alzheimer’s
The presence of radiolabeled metabolites can result in diffi-
culties in quantification of the PET data. Moreover, the low
11
11
brain uptake of (R)-[ C]verapamil ([ C]-1) limits its use. Espe-
[3–5]
disease and Parkinson’s disease.
Many studies have demon-
cially, if there is overexpression of P-gp at the blood–brain bar-
[
12]
strated that P-gp and other ABC transporters such as breast
cancer resistance protein (BCRP) and multidrug resistance asso-
ciated protein 1 (MRP1) play a pivotal role in controlling b-
amyloid peptide levels in the brain to avoid formation of amy-
rier, the uptake of the tracer will be even lower than that in
normal tissue. To overcome these limitations, a strategy to use
a selective P-gp inhibitor could be an option. Therefore, several
labeled inhibitors have been evaluated for PET imaging of P-
11
11
11
gp expression, including [ C]laniquidar, [ C]tariquidar ([ C]-2),
1
1
18
18
[
a] Dr. M. Cantore, M. Benadiba, P. H. Elsinga, C. Kwizera, R. A. J. O. Dierckx,
G. Luurtsema
[ C]elacridar, [ F]fluoroethyltariquidar, [ F]fluoroethylelacridar,
18
[13]
and 1-[ F]fluoroelacridar (Figure 1).
Department of Nuclear Medicine and Molecular Imaging
University Medical Center Groningen, University of Groningen
Hanzeplein 1, 9713 GZ Groningen (The Netherlands)
E-mail: mariangelacantore@gmail.com
However, it was observed that these compounds, which act
as inhibitors at pharmacological dose, are at tracer concentra-
tions recognized by P-gp and BCRP as substrates, which results
[
13–15]
[
b] Dr. M. Cantore, N. A. Colabufo
in low brain uptake of the PET probes.
These findings en-
Biofordrug s.r.l., Spin-off dell’Università degli Studi di Bari “A. Moro”
via Orabona 4, 70125, Bari (Italy)
Fax: (+39)0805442231
courage the search and screening of more specific, novel ra-
diolabeled compounds that could image either P-gp or BCRP
[
16–18]
transporters.
[
c] N. A. Colabufo
Given that ABC transporters are flexible proteins that bind
to several distinct molecules, it is a challenge to find specific
Dipartimento di Farmacia-Scienze del Farmaco
Università degli Studi di Bari “A. Moro”, via Orabona 4, 70125, Bari (Italy)
ChemMedChem 2016, 11, 108 – 118
108
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 1. Radiolabeled P-gp/BCRP substrates/inhibitors.
radiolabeled compounds that are capable of measuring
a single transporter subtype in vivo. Colabufo and colleagues
tested several P-gp modulators that behave as substrates, in-
hibitors, or even inducers depending on slight changes in their
The potency (EC ) of each P-gp-modulating agent was de-
50
termined in Madin–Darby canine kidney cells, stably transfect-
ed for P-gp expression (MDCK-MDR1). The selectivity toward
other ABC transporters could be evaluated in the same cell
line overexpressing either MRP1 (MDCK-MRP1) or BCRP
(MDCK-BCRP). This stepwise investigation (i.e., B!A/A!B
ratio ! ATP depletion ! affinity for P-gp ! selectivity for P-
gp) represents a suitable screening of P-gp candidates for radi-
olabeling.
[19–22]
molecular structures.
In a preclinical study, two new PET
11
11
probes were developed: a P-gp inhibitor [ C]MC18 ([ C]-3),
which showed specific binding in the brain and several periph-
eral target organs and is, therefore, useful to assess P-gp ex-
11
11
pression, and [ C]MC266 ([ C]-4), a P-gp substrate with higher
11
baseline uptake than [ C]-1 suitable for measurement of P-gp
On the basis of our previous studies, a new class of com-
[
18]
[24]
activity.
In
a
recent PET study, the new compound
pounds, cycloisosteres of MC113, was developed. MC150 (5)
1
1
[17]
[
C]MC113 was tested in mice. However, although this tracer
and the corresponding fluoromethyl and fluoroethyl deriva-
11
showed a higher target organ uptake than [ C]-4, it failed to
provide a P-gp-specific signal both at the mouse blood–brain
barrier and in a tumor model. Probably, the tracer has insuffi-
cient binding affinity to P-gp.
tives (i.e., MC236, 6, and MC211, 7) were selected as potential
11
18
tracers suitable for C and F labeling. The insertion of a fluo-
roalkyl moiety can change the biological and metabolic profile
of a compound. It is known that slight changes can lead to dif-
ferent P-gp activities. The effect of structural changes of the la-
beled alkyl moieties of these tracers was evaluated by in vitro
assays. The mechanism of interaction of each compound with
P-gp and its metabolic profile were determined. In the present
[23]
On the basis of Polli’s classification, it is possible to dis-
criminate between P-gp inhibitors, substrates, and substrate
[
21]
modulators. Therefore, we determined the apparent permea-
bility (Papp) for each developed compound in the Caco-2 cell
line. In this model system, the basolateral-to-apical (B!A) and
apical-to-basolateral (A!B) fluxes for each ligand were evalu-
ated. The B!A flux represents the contribution of passive dif-
fusion, whereas the A!B flux represents P-gp-mediated active
transport. The B!A/A!B ratio is useful to distinguish trans-
ported drugs from compounds that are not transported. Com-
pounds displaying B!A/A!B >2 are considered P-gp sub-
strates, whereas drugs showing B!A/A!B <2 are not trans-
ported by P-gp. Compounds that were not transported by P-
gp did not deplete ATP in this cell line, whereas compounds
transported by P-gp depleted ATP in a dose- and time-depen-
dent manner. These assays can discriminate between a P-gp in-
11
18
study, the radiosynthesis and in vitro evaluation of [ C]-5, [ F]-
18
6, and [ F]-7 are presented: our purpose was the development
of novel tracers for imaging P-gp activity and expression.
Results and Discussion
Compounds 5 and 7 presented in this work were synthesized
in good yields (40–50%), whereas compound 6 was synthe-
sized in low yield (less than 5%) as a result of difficulties in in-
serting the small fluoromethyl chain. For this reason, com-
pound 6 was chemically characterized but not tested in vitro
(Scheme 1).
[21,23]
11
18
18
hibitor and a P-gp substrate.
Compounds [ C]-5, [ F]-7, and [ F]-6 (Scheme 2) were pre-
11
pared in good radiochemical yields, >40% yield for [ C]-5 and
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
109
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
different time points. Both compounds were stable during this
interval. After 120 min of microsomes incubation with 5, 53%
of the total mass still represented parent compound, and the
contribution of O-demethylation metabolites was 15%. Deme-
thylation reactions occur within 15 min. In a metabolic study
with compound 1, 45% parent remained after 30 min, togeth-
er with 11 % of demethylation products. In contrast, 5 and 7
showed higher metabolic stability than compound 1. After
1
20 min of microsomial incubation with 5, 53% of the parent
remained, with 15% O-demethylated metabolites. For com-
pound 7, 57% of the parent compound remained after
1
20 min, together with 15, 12, and 3% products of oxidative
defluorination, demethylation, and F-dealkylation, respectively
Figure 2). Loss of the fluoroethyl moiety started after 45 min.
(
For both compounds, cleavage of an NÀC bond caused forma-
tion of 6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline, which
was identified by MS as a primary metabolite. Metabolic stud-
ies in microsomes provide additional information about the
stability of the moiety suitable to labeling in candidates radioli-
gands. These findings indicate that from a metabolic point of
view, 5 and 7 might be better PET radiotracers candidates than
compound 1.
Scheme 1. Synthesis of MC150 (5), MC211 (7), MC236 (6), and precursor 12.
Reagents and conditions: a) MeI, Cs CO , DMF, 30 min, RT (46%); b) 7n
NH OH/MeOH, 608C, 6 h (90%); c) 1,3-dichloroacetone, 1508C, 5 h (54%);
d) 6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline, Na CO , DMF, 1508C, over-
night (82%); e) NaH, DMF, fluoromethyl tosylate or fluoroethyl tosylate,
508C, 24 h (2% for 6, 38% for 7).
2
3
4
2
3
1
Compounds 5 and 7 were investigated for their
mechanism of interaction with P-gp and their selec-
tivity toward the other efflux pumps MRP1 and
[
24]
BCRP. Because of the low stability of the fluoro-
methyl moiety under the standard experimental
conditions, in in vitro biological assays, compound
18
[
F]-6 was only investigated in the following func-
tional assays: 1) apparent permeability determina-
tion (Papp) in Caco-2 cell monolayer; 2) inhibitory ac-
tivity of the acetoxymethyl esters of calcein-AM
transport in MDCK cells overexpressing each trans-
porter (i.e., MDCK-MDR1, MDCK-BCRP, MDCK-MRP1);
3
) ATP depletion in MDCK cells. For compound 5,
the transport ratio through Caco-2 monolayers in
the basolateral!apical and apical!basolateral di-
À1
rections (P
B!AA !B) was 10, and the com-
app
pound did not activate ATPase activity within the
monolayer. Compound 5 showed moderate activity
toward P-gp (EC =3.1 mm) and BCRP (EC =2 mm)
50
50
in the calcein-AM assay and very low activity toward
MRP1 (EC >50 mm). For compound 7, the transport
ratio through Caco-2 monolayers in the basolater-
50
1
1
18
18
Scheme 2. Radiosynthesis of [ C]-5, [ F]-6, and [ F]-7. Reagents and conditions:
1
8
11
a) FCH
(
2
OTf, NaH, DMF, 10 min, 1258C (3–10%); b) CH
40%); c) FCH CH OTs, NaH, DMF, 15 min, 1258C (3–10%).
3
I, DMSO, KOH, 1258C, 1 min
18
al!apical and apical!basolateral directions (P
app
2 2
B!A/A!B) was 2.8, and the compound did not ac-
tivate ATPase activity within the monolayer. Com-
18
18
3
–10% yield for [ F]-7 and [ F]-6, with high specific activity
>50–100 GBqmmol ). For the preparation of [ C]-5, the use
of [ C]CH I as a methylating agent led to a successful synthe-
pound 7 showed high affinity for P-gp (EC =0.8 mm) and very
50
À1
11
(
low activity toward BCRP in the calcein-AM assay (EC >
50
11
50 mm), and it was inactive toward MRP1 (EC >100 mm). On
3
50
18
18
sis, whereas for [ F]-6, the [ F]fluorobromomethane synthetic
route was not reliable and was characterized by a low yield.
Compounds 5 and 7 were tested for their metabolic stability
in human liver microsomes in combination with UPLC–MS-
QTOF (ultraperformance liquid chromatography quadrupole
time-of-flight mass spectrometry). The metabolic profile of
each compound was monitored for 2 h and was evaluated at
the basis of these data, compound 5 was classified, taking into
[
23]
account the classification reported by Polli et al., as a dual P-
gp/BCRP-transported substrate, and compound 7 could be es-
timated as a selective P-gp inhibitor, although its B!A/A!B
ratio is borderline for substrate/inhibitor activity.
All radioligands were evaluated in the P-gp-overexpressing
cell line Colo320. After separation of subcellular fractions, the
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
110
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 2. Metabolic pathways for compound 7 in human liver microsomes.
uptake/binding of each tracer in each specific fraction was de-
termined. It is known that P-gp can also be highly expressed in
the membranes of vesicles that are located in the cytoplasm.
less transported (Figure 5a). Interestingly, after pretreatment
with P-gp modulators 1 and 13 and P-gp inhibitor 3 (Fig-
11
18
A very low cellular accumulation of [ C]-5 and [ F]-6 was
1
8
found, significantly lower than that of [ F]-7 (Figure 3a,b).
11
18
These data demonstrate that [ C]-5 and [ F]-6 are potent P-
18
gp substrates showing strong efflux. In contrast, [ F]-7
showed more specific binding to P-gp in both the cytoplasmic
and membrane fractions (Figure 3b).
To confirm or refute that these compounds are selectively
transported or effluxed by P-gp, a fractionation experiment
was performed in the breast cancer cell line MDA-MB-231,
which does not express P-gp but has overexpression of BCRP
[
25]
18
(
Figure 4a,b). The results demonstrate that [ F]-7 is signifi-
cantly less accumulated in the cytoplasmic fraction (Figure 4a)
and more accumulated in the membrane fraction (Figure 4b)
11
of this cell line relative to [ C]-5, which is transported by BCRP.
This is in agreement with the previous results that show that 5
is a potent substrate for both P-gp and BCRP, whereas 7 is
a compound with high affinity for P-gp and is thought to be
18
an inhibitor. Interestingly, the results demonstrate that [ F]-7
shows significant binding even if only a very tiny amount of
the P-gp transporter is present (see the percentage of injected
dose in Figure 3 and compare this with Figure 4 relating to
MDA-MB). Thus, in cells that do not express a high amount of
P-gp, more cytoplasmic accumulation of the high-affinity sub-
11
11
18
strate [ C]-5 might be found.
Figure 3. a) Cytoplasmic and b) membrane accumulation of [ C]-5, [ F]-7,
1
8
18
and [ F]-6 in the P-gp-overexpressing Colo320 cell line. Note that [ F]-7 is
less transported or effluxed outside the cells with a more specific binding to
P-gp in both cytoplasmic and membrane fractions. Data are the meanÆ-
SEM, n=6. Student t-test was performed; p<0.05 was considered statistical-
ly significant.
In the transwell permeability assay, the transport of each
tracer in Colo320 cells was evaluated. In accordance with the
results mentioned above, we found that all radiolabeled com-
1
8
pounds were effluxed to the apical side; however, [ F]-7 was
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
111
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
18
transport proteins in cells that could bind [ F]-6 efficaciously if
P-gp is modulated. Unfortunately, as reported above, unla-
beled 6 is not available, and therefore, additional investiga-
tions concerning its specificity could not be performed. Com-
11
pound [ C]-1 was also tested in this experiment (Figure 5e). It
11
was transported more than [ C]-5, but its permeation was in-
creased only 5-fold by 3 and only 3.5-fold by 13. For all the
compounds, increased permeation after modulator pretreat-
ment could be attributed to competition between the com-
pounds and the modulators for the same binding site.
11
An additional experiment with [ C]-1 as a positive control
was performed (Figure 6). In accordance with our speculation,
no competition was observed between P-gp modulators 1 and
1
8
1
3 and radiolabeled compound [ F]-7. However, both modula-
11
11
tors interacted with [ C]-5 and [ C]-1 (Figure 6). This interac-
11
tion suggests that [ C]-5 binds to the same binding site as
and 13. In contrast, [ F]-7 accumulated more in both frac-
18
1
tions, which suggests that this radiotracer could dislocate the
P-gp modulators or bind to a different site on the P-gp mole-
18
cule. Anyway, [ F]-7 seems to have greater affinity for P-gp
and, therefore, presented a strong and noncompetitive bind-
ing property, as it is highly accumulated in the cytosol and in
the membrane fractions (Figure 6a,b) even if pretreated with
P-gp modulator 1 or 13. Ligands 1 and 13 competed with
1
1
18
Figure 4. a) Cytoplasmic and b) membrane accumulation of [ C]-5 and [ F]-
7
in the MDA-MB breast cancer cell line. Data are the meanÆSEM, n=6.
Student t-test was performed; p<0.05 was considered statistically signifi-
cant.
18
[
F]-6 in both fractions, although 13 is reported to be a P-gp
ure 5b), it was found that only inhibitor 3 significantly in-
modulator rather than a P-gp substrate. In addition, as the
nonradioactive drugs are added 30 min before the radiotracer,
18
creased [ F]-7 permeation to the basolateral side. In contrast,
11
11
18
all modulators increased the permeation of [ C]-5 approxi-
it is also possible that [ C]-5 and [ F]-6 cannot displace li-
gands if they are already attached to the same binding site of
the P-gp molecule (resulting in less accumulation of the radio-
tracer). Interestingly, previously characterized P-gp inhibitor
1
8
mately sixfold (Figure 5c). Compound [ F]-6 presented a differ-
ent profile, as pretreatment with compound 1 increased its
permeation 35-fold and pretreatment with compound 3 in-
creased it only 1.3-fold, whereas pretreatment with 13 in-
creased its efflux significantly. This discrepancy could be ex-
plained by the assumption that there are many constitutive
[
26]
3
increased the permeation of all radiolabeled compounds
18
to the basal side. However, for inhibitor [ F]-7 we can con-
clude that increased permeation to the basal side is due to
1
1
18
Figure 5. a) Transwell permeability assay in the Colo320 cell line showing a higher transport activity of [ C]-5 to the apical side than the inhibitor [ F]-7.
1
8
11
18
Transwell permeability assays in the Colo320 cell line after pretreatment with MC18 (3), verapamil (1), and cyclosporine A (13) for b) [ F]-7, c) [ C]-5, d) [ F]-
1
1
6
, and e) [ C]-1. A/B: apical/basolateral ratio; data are the meanÆSEM, n=3.
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
112
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
specific activity and in acceptable radiochemical yields. The
preliminary in vitro results of cold compounds were confirmed
at the tracer level, which confirms that small changes in the
molecular structures could lead to different P-gp and BCRP ac-
tivities. These data can lead to better understanding of the
main structural requirements for the development of novel P-
11
gp tracers. Compound [ C]-5 represents a new substrate, with
1
1
11
P-gp activity similar to that of (R)-[ C]verapamil ([ C]-1) and
11
improved metabolic stability. However, with respect to [ C]-1,
11
ligand [ C]-5 is more sensitive in terms of cell uptake in the
presence of inhibitor and modulator than verapamil, and this
aspect is important because cell radiotracer accumulation is
18
employed to quantify P-gp activity. On the other hand, [ F]-7
18
represents a new P-gp inhibitor. For [ F]-6, it is important to
determine its mechanism of interaction with P-gp, although
several pharmacokinetic limitations discourage these additional
studies.
Experimental Section
General procedures
Column chromatography was performed with 1:30 Merck silica gel
6
0 A8 (63–200 mm) as the stationary phase. Melting points were de-
termined in open capillaries with a Gallenkamp electrothermal ap-
1
paratus. H NMR spectra were recorded in CDCl at 300 MHz with
3
a Varian Mercury-VX spectrometer. All spectra were recorded on
the free bases. All chemical shift values are reported in ppm (d).
Recording of mass spectra was done with an HP6890–5973 MSD
gas chromatograph/mass spectrometer; only significant m/z peaks,
with their percentages of relative intensity in parentheses, are re-
ported. All spectra were in accordance with the assigned struc-
tures. ESI-MS analyses were performed with an Agilent 1100 LC–MS
Dtrap system VL. Purity of final compounds was established by
combustion analysis of the corresponding hydrochloride salts,
which confirmed a purity ꢀ95%.
For radiolabeled compounds, detection on TLC was performed
with Cyclone phosphor storage screens (multisensitive, Packard,
PerkinElmer Life and Analytica Science). These screens were ex-
posed to the TLC strips and subsequently read out by using a Cy-
clone phosphor storage imager (PerkinElmer) and were analyzed
with OptiQuant software. Radiolabeled compounds were also iden-
tified by analysis by using a Waters UPLC Acquity H class with
a BEH-C18 analytical column (2.1”50 mm, 1.7 mm). A gradient of
1
8
11
18
Figure 6. Competition assay in the Colo320 cell line for [ F]-7, [ C]-5, [ F]-6,
11
and [ C]-1. a) Cytoplasmic and b) membrane fraction obtained from the P-
gp-overexpressing cell line Colo320 after pretreatment with P-gp modula-
tors 1 (dark grey) and 13 (light grey). Data are corrected for the values ob-
tained from the controls. %ID: percentage of the injected dose. Data are the
meanÆSEM, n=6. Student t-test was performed; p<0.05 was considered
statistically significant.
a competition between tracer and inhibitor drug for the same
binding site. This result also confirmed our previous hypothesis
suggesting that the P-gp modulators verapamil and cyclospor-
1
8
in bind to a site different from the binding site of [ F]-7 (Fig-
ure 6b).
11
In summary, methyl derivative [ C]-5 represents a new meta-
bolically stable P-gp/BCRP substrate, whereas insertion of a flu-
1
8
oroethyl moiety instead of a methyl group results in [ F]-7,
a P-gp inhibitor. Insertion of a fluoromethyl moiety instead of
0
.1% formic acid/acetonitrile was used as the mobile phase at
À1
1
8
a flow rate of 0.6 mLmin (formic acid from 98 to 2% in 16 min)
and electrospray ionization (ESI ) mass spectra were acquired at
a fluoroethyl group leads to compound [ F]-6, which shows
a different profile: it binds to the same site as 1, but in the
presence of inhibitor 3, its permeation increases a few fold,
whereas in the presence of 13 its efflux increases.
+
unit resolution from m/z=200 to 1200.
HPLC purifications and analysis were performed with an Elite LaCh-
rom VWR Hitachi L-2130 pump system connected to a UV spec-
trometer (Elite LaChrom VWR Hitachi l-2400 UV detector) set at
a wavelength of 260 nm and a radiation detector (Bicron). Semipre-
parative HPLC was performed by using a Phenomenex Luna C18-
column (10 mm”250 mm, 5 mm), with sodium acetate/acetonitrile
Conclusions
1
1
In this work, three new compounds, 2-[2-(2-methyl-( C)-5-me-
thoxyphenyl) oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetrahy-
À1
(6:4) as the mobile phase at a flow rate of 4 mLmin . Quality con-
11
18
droisoquinoline ([ C]-5), 2-[2-(2-fluoromethyl-( F)-5-methoxy-
trol was performed on a reverse-phase Symmetric Shield-C18
column (4.6 mm”250 mm, 5 mm) with sodium acetate/acetonitrile
(6:4) as the mobile phase at a flow rate of 1 mLmin .
phenyl)oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetra-hydroiso-
À1
18
18
quinoline ([ F]-6), and 2-[2-(2-fluoroethyl-( F)-5-methoxyphe-
nyl)oxazol-4-ylmethyl]-6,7-dimethoxy-1,2,3,4-tetrahydroisoqui-
All products were formulated with the following procedure. The
mixture was passed over an Oasis HLB Sep-Pak to remove acetoni-
18
noline ([ F]-7), were presented. All were synthesized with high
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
113
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
trile, and the cartridge was washed with water (2”8 mL), and then
the product was eluted from the Oasis HLB Sep-Pak with sterile
ethanol (1 mL). The formulated product was suitable to perform all
in vitro experiments. Radioactivity measurements during cell stud-
ies were done by using an automated gamma counter (LKBG-Com-
pugamma CS 1282, Wallac, Waltham, USA).
of growth were used. Cells were pre-incubated for 30 min with dis-
tinct P-gp modulators together with or without ATP, followed by
30 min tracer incubation.
Synthesis
Nonlabeled chemistry
Materials
Compounds 5, 6, and 7 and precursor 12 were synthesized as re-
ported in Scheme 1. Compound 5 was synthesized as previously
reported, and spectral data were identical to those previously de-
Commercially available chemicals were purchased from Sigma–Al-
drich. Verapamil was purchased by Sigma–Aldrich; tariquidar was
purchased by API Services, Inc.; cyclosporine A (Sandimmune,
[23]
scribed. Intermediates 10a and 10b were prepared by ammo-
À1
5
0 mgmL solution in polyoxyethylated Ricinus oil) was a product
nolysis of methyl esters 9a and 9b. Amides 10a and 10b were cy-
clized in the presence of 1,3-dichloroacetone to give oxazoles 11 a
and 11 b, which were condensed with 6,7-dimethoxy-1,2,3,4-tetra-
hydroisoquinoline to obtain precursor 12 and 5. Compound 7 was
synthesized from phenol 12 and fluoroethyl tosylate, which was
previously prepared by reaction of fluoroethanol and p-toluensul-
fonyl chloride. Compound 6 was prepared from precursor 5 by
reaction with fluoromethyl tosylate previously synthesized from
bis(tosylate) and silver p-toluenesulfonate.
of Sandoz, adenosine 5’-triphosphate (ATP) was purchased from
Sigma–Aldrich. Stock solutions of 10 mm of these compounds
were prepared fresh prior to the experiment. Subcellular fraction
buffer was prepared by the combination of the following reagents:
2
50 mm sucrose, 20 mm HEPES (2-[4-(2-hydroxyethyl)piperazin-1-
yl]ethanesulfonic acid; pH 7.4; Life Technologies), 10 mm KCl
Merck), 1.5 mm MgCl (Merck), 1 mm EDTA (ethylenediaminetetra-
(
2
acetic acid; Merck), 1 mm EGTA (ethylene glycol tetraacetic acid;
Sigma–Aldrich), DTT (dithiothreitol; Sigma–Aldrich), and protease
inhibitor cocktail (Sigma–Aldrich).
Methyl 2-hydroxy-5-methoxybenzoate (9a): 2-Hydroxy-5-methox-
ybenzoic acid (8a, 1 mmol) was dissolved in DMF (5 mL). Cesium
carbonate (0.8 mmol) was added, and the mixture was stirred at
QMA-light Sep-Pak cartridges were purchased from Waters and
were conditioned with 1.4% aq. sodium hydrogen carbonate
0
8C for 10 min. CH I (1.1 mmol) was added dropwise over 5 min,
3
(
5 mL) and water (100 mL) prior to use. Silica Sep-Pak cartridges,
and the mixture was warmed to RT. After 30 min, water (30 mL)
was added, and the aqueous layer was extracted with EtOAc (3”
Sep-Pak Alumina N cartridge, and Oasis HLB Sep-Pak were pur-
chased from Waters and were used as received without any pre-
conditioning. A silver triflate column was prepared as described
2
0 mL) and Et O (3”20 mL). The combined organic layer was
2
washed with 10% NaHCO (50 mL) and a saturated aqueous solu-
[27]
11
11
3
previously. The yield of conversion of [ C]CH I into [ C]CH tri-
3
3
tion of NaCl (3”30 mL), dried (Na SO ), and concentrated under re-
2
4
flate was calculated with a radioassay by employing 4-(4-nitroben-
duced pressure to give an oily residue. The crude compound was
[28]
zyl)pyridine. The column was heated to 1908C under He flow
purified by column chromatography (CHCl /EtOAc=1:1) to give
À1
3
(
110 mLmin ) before starting the synthesis.
+
a brown oil (46%). GC–MS: m/z (%): 182 (42) [M] , 150 (100), 107
(
29).
Cell lines
2-Hydroxy-5-methoxybenzamide (10a): Methyl ester 9a (4 mmol)
was dissolved in methanolic ammonia solution (15 mL, 7n), and
the mixture was warmed at 608C for 5 h. The solution was cooled
and concentrated to dryness. The crude product was purified by
Cell culture reagents were purchased from Celbio s.r.l. (Milan, Italy).
CulturePlate 96-well plates were purchased from PerkinElmer Life
Science; calcein-AM and MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-di-
phenyltetrazoliumbromide) were obtained from Sigma–Aldrich
column chromatography (CHCl /EtOAc=1:1) to give a white solid
3
1
(
90%); mp: 148–1498C. H NMR: d=3.79 (s, 3H, CH ), 5.96 (br, 2H,
3
(
Milan, Italy). MDCK-MDR1, MDCK-BCRP, and MDCK-MRP1 cell lines
NH , exchange with D O), 6.93–7.08 (m, 3H,) 11.59 ppm (br, 1H,
2
2
were a gift of Professor P. Borst, NKI-AVL Institute, Amsterdam, The
Netherlands. All MDCK cells were grown in Dulbecco’s modified
Eagle’s medium (DMEM) high glucose supplemented with 10%
+
OH, exchange with D O). GC–MS: m/z (%): 167 (37) [M] , 150 (100),
2
1
07 (40), 79 (26).
À1
fetal bovine serum, 2 mm glutamine, 100 UmL penicillin, and
2-[4-(Chloromethyl)oxazol-2-yl]-4-methoxyphenol (11a): Amide
10a (1 mmol) was condensed with 1,3-dichloroacetone (2 mmol)
À1
1
00 gmL streptomycin in a humidified incubator at 378C with
a 5% CO atmosphere. Caco-2 cells were a gift of Dr. Aldo Cavallini
with heating at 1508C for 5 h. After cooling to RT, H O (10 mL) and
2
2
and Dr. Caterina Messa from the Laboratory of Biochemistry, Na-
tional Institute for Digestive Diseases, “S. de Bellis”, Bari, Italy.
EtOAc (10 mL) were added. The aqueous phase was extracted with
EtOAc (3”50 mL), and the combined organic layer was dried
(
Na SO ) and concentrated. The residue was purified by column
2 4
P-gp-overexpressing Colo320 colon carcinoma and the P-gp nonin-
duced MDA-MB breast cancer cell lines were kindly provided by
the Medical Oncology Department of the UMCG, and stocks were
maintained frozen (liquid nitrogen) in Roswell Park Memorial Insti-
tute medium (RPMI) and DMEM respectively, supplemented with
chromatography (hexane/EtOAc=9:1) to give a yellow solid (54%).
1
H NMR: d=3.82 (s, 3H, CH ), 4.56 (s, 1H, CH ), 4.57 (s, 1H, CH ),
3
2
2
6
.98–7.02 (m, 2H), 7.25 (s, 1H), 7.69 (s, 1H), 10.45 ppm (br, 1H, OH,
+
exchange with D O); GC–MS: m/z (%): 241 (32) [M+2] , 239 (94)
2
+
[M] , 224 (100), 176 (26).
1
0% fetal calf serum and 10% sodium dodecyl sulfate (SDS). Stock
cells were grown in RPMI or DMEM plus 1% glutamine containing
0% fetal calf serum without antibiotics. From a more or less con-
2-[2-(2-Hydroxy-5-methoxyphenyl)oxazol-4-ylmethyl]-6,7-dime-
thoxy-1,2,3,4-tetrahydroisoquinoline (12): A mixture of 4-chloro-
methyl-2-aryloxazole 11 a (0.50 mmol), 6,7-dimethoxytetrahydroiso-
quinoline (1.2 mmol), and Na CO (0.50 mmol) in DMF (5 mL) was
stirred overnight. The solvent was evaporated, and the crude prod-
uct was washed with H O (2”20 mL) and extracted with CHCl
1
2
fluent 25 cm flask, cells were harvested by trypsinization (trypsin
.025%) and then counted by using an improved Neubauer count-
ing chamber. Cells were seeded into 24-well plates at a density of
0
2
3
4
7
5
”10 cells per well and incubated in a humidified atmosphere of
2
3
% CO , 95% air at 378C. After 48 h, cells in the exponential phase
(30 mL). The organic layer was dried (Na SO ) and concentrated
2 4
2
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
114
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
under reduced pressure. The crude product was purified by
2-[2-(2-Fluoromethyl-5-methoxyphenyl)oxazol-4-ylmethyl]-6,7-
dimethoxy-1,2,3,4-tetrahydroisoquinoline (6): A suspension of
NaH (60%, mmol) in dry DMF (3 mL) was stirred at RT for 10 min. A
solution of phenol 12 (1 mmol) in DMF (1 mL) was added, and the
mixture was stirred for 1 h. A solution of 2-fluoromethyl tosylate
(1.1 mmol) in DMF (1 mL) was added, and the mixture was stirred
column chromatography (CHCl /MeOH=19:1) and recrystallized
3
1
(
MeOH/Et O). Yield: 82%. H NMR (CDCl ): d=2.78–2.81 (m, 4H,
2
3
NCH CH ), 3.74 (s, 2H, CH NCH ), 3.75 (s, 2H, CH NCH ), 3.81 (s, 3H,
2
2
2
2
2
2
CH ), 3.82 (s, 6H, CH ), 3.82 (s, 3H, CH ), 6.59 (s, 1H), 6.51 (s, 1H),
3
3
3
6
.96–6.98 (m, 2H), 7.30 (s, 1H), 7.64 (s, 1H), 10.75 ppm (s, br, 1H,
+
+
OH, exchange with D O); MS (ESI ): m/z: 419 [M+Na] ; MS–MS
for 24 h. H O was added until effervescence ceased. The solvent
2
2
+
(
ESI ): m/z (%): 404 (4), 226 (100); Anal. calcd for
was evaporated, and the residue was partitioned between H O
2
C H N O HCl·0.5H O: C 66.65, H 6.10, N 7.07, found: C 66.15, H
(20 mL) and CHCl (20 mL). The organic phase was separated, and
22
24
2
5
2
3
6
.02, N 6.98.
the aqueous phase was extracted with CHCl (3”50 mL). The com-
3
bined organic layer was dried (Na SO ) and concentrated. The resi-
2
4
2
-Fluoroethyltosylate: A solution of 2-fluoroethanol (2 mmol) and
due was purified on silica gel column chromatography (CHCl /
3
p-toluenesulfonyl chloride (1.1 mmol) in 5m NaOH (1.6 mmol) was
stirred at RT for 24 h. The mixture was diluted with CH Cl , and the
organic phase was washed with 10% NaOH. The organic layer was
dried (Na SO ) and concentrated under reduced pressure. The
crude product was purified by column chromatography (CH Cl ) to
give a colorless oil (92%). H NMR: d=2.45 (s, 3H, CH ), 4.21 (t, 1H,
CH ), 4.30 (t, 1H, CH ), 4.49 (t, 1H, CH ), 4.64 (t, 1H, CH ), 7.34 (d,
J=8 Hz, 2H), 7.80 ppm (d, J=8 Hz, 2H); MS (ESI ): m/z: 241 [M+
Na] ; MS–MS (ESI+): m/z (%): 241 (74), 97 (100).
+
+
MeOH=19:1). Yield: 2%; MS (ESI ): m/z: 451 [M+Na] ; MS–MS
ESI+): m/z (%): 431 (4), 409 (100).
2
2
(
2
4
Radiolabeled chemistry
2
2
1
3
11
11
[
C]-1: [ C]-1 was produced from desmethyl-1 as previously de-
2
2
2
2
[27]
scribed. The radiochemical yield was ꢀ60% end of bombard-
+
11
ment (EOB) based on [ C]methyl triflate), radiochemical purity
+
À1
ꢀ
[
99%, and the specific activity ꢀ100 GBqmmol .
C]-5: [ C]Methyl iodide was added to a solution of DMSO
2
-[2-(2-Fluoroethyl-5-methoxyphenyl)oxazol-4-ylmethyl]-6,7-di-
methoxy-1,2,3,4-tetrahydroisoquinoline (7): A suspension of NaH
60%, 1.2 mmol) in dry DMF (3 mL) was stirred at RT for 10 min. A
11
11
(
(
300 mL) containing precursor 12 (1 mg, 2.2 mmol) and KOH
14 mg). After trapping [ C]methyl iodide, the temperature was
(
11
solution of phenol 3a (1 mmol) in DMF (1 mL) was added, and the
solution was stirred for 1 h. A solution of 2-fluoroethyl tosylate
raised to 1258C for 1 min, and after this time the mixture was
quenched with 1m HCl (1 mL). The mixture was purified and for-
mulated as reported in the section General Procedures. The radio-
(
1.1 mmol) in DMF (1 mL) was added, and the mixture was stirred
for 4 h. Water was added until effervescence ceased. The solvent
was evaporated, and the residue was partitioned between H O
11
chemical yield was ꢀ40% (EOB based on [ C]methyl iodide), radi-
2
À1
ochemical purity ꢀ99%, and the specific activity ꢀ100 GBqmmol
(
20 mL) and CHCl (20 mL). The organic phase was separated, and
3
(
EOB values).
the aqueous phase was extracted with CHCl (3”50 mL). The com-
bined organic fraction was dried (Na SO ) and concentrated. The
residue was purified by column chromatography (CHCl /MeOH=
3
1
8
18
Drying procedure for [ F]fluoride: Aqueous [ F]fluoride was pro-
2
4
18
18
duced by the O (p,n) F nuclear reaction by irradiation of
[ O]water with a Scanditronix MC-17 cyclotron. The [ F]fluoride
solution was passed through an activated Sep-Pak Light Accell
plus QMA anion-exchange cartridge (Waters) to recover [ O]water.
[ F]fluoride was eluted from the cartridge with K CO (5 mgmL ,
2 3
3
1
18
18
1
2
3
9:1) and recrystallized (MeOH/Et O). Yield: 38%. H NMR: d=
.88–3.19 (m, 4H, NCH CH ), 3.80–3.98 (m, 4H, CH NCH ), 3.81 (s,
2
2
2
2
2
1
8
H, CH ), 3.83 (s, 3H, CH ), 3.84 (s, 3H, CH ), 4.21–4.33 (m, 2H,
3
3
3
18
À1
CH CH F), 4.68–4.86 (m, 2H, CH CH F), 6.52 (s, 1H), 6.60 (s, 1H),
2
2
2
2
+
6
4
1
.97–6.98 (m, 2H), 7.49 (s, 1H), 8.02 ppm (s, 1H); MS (ESI ): m/z:
43 [M+H] ; MS–MS (ESI ): m/z (%): 250 (100), 222 (44), 195 (47),
49 (13); Anal. calcd for C H FN O HCl: C 65.15, H 6.15, N 6.33,
1 mL) and was collected in a vial containing Kryptofix 2.2.2
(15 mg). Acetonitrile (dried on molecular sieves, 1 mL) was added
to this solution, and the solvents were evaporated at 1308C. The
+
+
2
4
27
2
5
18
found: C 64.95, H 5.98, N 6.21.
[ F]KF/Kryptofix complex was dried three times by the addition of
acetonitrile (0.5 mL), followed by evaporation of the solvent.
Bis(tosyloxy)methane: A solution of silver p-toluenesulfonate (0.05
mol) and methyl iodide (0.023 mol) in acetonitrile (50 mL) was
heated at reflux for 24 h. The mixture was filtered, and the volatile
solvent was removed under reduced pressure. The solid product
was dissolved in warm 1,2-dichloroethane (150 mL) and then fil-
tered to remove the excess amount of silver methanesulfonate.
The solvent was removed by distillation to give a solid that was re-
crystallized (absolute ethanol). Yield: 87%; mp: 116–1178C; MS
1
8
18
[ F]Fluoromethyl triflate: The synthesis of [ F]bromofluoro-
methane was performed in an automated synthesis apparatus
(Fluoro Methylation Unit, model FMU-101, Veenstra) A solution of
dibromomethane (500 mL) in acetonitrile (700 mL) was added to the
18
dry [ F]KF/Kryptofix complex. The mixture was heated at 115 8C for
5 min, and after the reaction, formed [ F]CH BrF was passed
through four Sep-Pak silica cartridges connected in series by He
18
2
+
+
+
À1
(
(
ESI ): m/z: 379 [M+Na] ; MS–MS (ESI ): m/z (%): 349 (100), 155
2).
flow (100 mLmin ) to separate the product from CH
2
Br
. Once
2
18
[ F]CH
BrF started to elute from the Sep-Pak, it was allowed to
2
flow through the heated (1908C) AgOTf column to convert it into
Methylene fluorotosylate: A solution of tetrabutylammonium
fluoride (1.1 mmol in 2 mL acetonitrile) was added to a solution of
bis(tosyloxy)methane (1 mmol) in dry acetonitrile (10 mL), and the
mixture was stirred at reflux. After 2 h, the solvent was removed
under reduced pressure, and the residue was dissolved in EtOAc
18
[
[
F]fluoromethyl triflate (decay-corrected radiochemical yield from
F]fluoride was 35–40%). [ F]Fluoromethyl triflate was trapped at
18
18
0
8C in a second reaction vessel containing the precursor for the
18
synthesis of [ F]-6.
1
8
(
30 mL) and washed with H O (30 mL). The organic solution was
[ F]-6: Precursor 12 (0.5 mg) was treated with NaH (5 mg) in dry
2
18
dried (MgSO ), filtered, and concentrated under reduced pressure.
DMF (0.2 mL). [ F]Fluoromethyl triflate was trapped in this solution
4
18
The residue was purified by column chromatography (hexane/
at 08C. After trapping of [ F]fluoromethyl triflate, the reaction
vessel was heated at 1258C for 10 min, and after cooling, the mix-
+
EtOAc=5:2). Yield: 31%. GC–MS: m/z (%): 204 (28) [M] , 155 (49),
À1
9
1 (100).
ture was diluted to 1 mL with 0.1 molL acetonitrile/sodium ace-
tate and filtered with HV filter (0.45 mm). The product was purified
and formulated as reported in the General Procedures. The radiola-
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
115
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
beled compound was obtained in 3–10% radiochemical yield from
Calcein-AM experiment: These experiments were performed as de-
1
8
[29]
[30]
[
[
(
F]fluoride (decay corrected) in 80–90 min. The specific activity of
scribed by Feng et al. and Rautio et al. with minor modifica-
tions. Madin–Darby canine kidney (MDCK) cells overexpressing
each transporter (i.e., MDCK-MDR1, MDCK-BCRP, and MDCK-MRP1)
1
8
À1
F]-6 was ꢀ50 GBqmmol and the radiochemical purity was 95%
EOS values)
were grown in DMEM high glucose supplemented with 10% fetal
1
8
[
F]Fluoroethyl tosylate: A solution of ethyleneglycol-1,2-ditosy-
À1
bovine serum, 2 mm glutamine, 100 UmL
penicillin, and
late (5 mg) in dry acetonitrile (0.5 mL) was added to the dried
À1
1
00 gmL streptomycin in a humidified incubator at 378C under
1
8
[
F]KF/Kryptofix complex, and the mixture was heated at 1258C
a 5% CO atmosphere. Each cell line (50000 cells per well) was
2
for 5 min. The product was purified by dissolving the product in
DMF (0.3 mL) and passing through a Sep-Pak Alumina N cartridge.
Kryptofix, K CO , and the excess amount of [ F]fluoride were re-
seeded into black CulturePlate 96-well plate with medium (100 mL)
and allowed to become confluent overnight. Each compound
1
8
2
3
(
100 mL) at various concentrations (1–100 mm) was solubilized in
tained on the cartridge, and the product was eluted with DMF
culture medium and was added to the monolayers. The 96-well
plate was incubated at 378C for 30 min. Calcein-AM was added in
of phosphate-buffered saline (PBS, 100 mL) to yield a final concen-
tration of 2.5 mm, and the plate was incubated for 30 min. Each
well was washed with ice-cold PBS (3”). Saline buffer was added
to each well, and the plate was read by a Victor 3 plate reader (Per-
kinElmer) at excitation and emission wavelengths of 485 and
535 nm, respectively. Under these experimental conditions, Calcein
cell accumulation in the absence and in the presence of the tested
compound was evaluated and the fluorescence basal level was es-
timated by untreated cells. In treated wells, the increase in fluores-
cence with respect to the basal level was measured. EC50 values
were determined by fitting the fluorescence increase percentage
versus log [dose].
(
[
0.3 mL). The procedure required approximately 50 min from
F]fluoride, decay corrected, and it provided [ F]fluoroethyl tosy-
1
8
18
late with an overall radiochemical yield of 20 to 40%.
1
8
[
(
F]-7: Precursor 12 (0.5 mg) and NaH (3–4 mg) dissolved in DMF
1
8
0.5 mL) were added to a solution of [ F]fluoroethyl tosylate in
DMF (0.5 mL). The mixture was heated for 15 min at 1258C and
after cooling down was diluted to 0.1 molL acetonitrile/sodium
acetate (1 mL) and then filtered through a HV filter (0.45 mm). The
product was purified and formulated as reported in the General
Procedures. The radiolabeled compound was obtained in 3–10%
radiochemical yield from [ F]fluoride (decay corrected) in 80–
9
the radiochemical purity was 95% (EOS values)
À1
1
8
1
8
À1
0 min. The specific activity of [ F]-7 was ꢀ100 GBqmmol and
ATPlite assay: The MDCK-MDR1 cells were seeded into a 96-well mi-
4
croplate in complete medium (100 mL) at a density of 2”10 cells
per well. The plate was incubated overnight under a humidified
Biological methods
[31]
atmosphere of 5% CO at 378C. The medium was removed and
2
In vitro experiments of nonradioactive 5 and 7
complete medium (100 mL) in the presence or absence of different
concentrations (1–100 mm) of each compound was added. The
plate was incubated for 2 h under a humidified atmosphere of 5%
Permeability experiments: Caco-2 cells were seeded onto a Millicell
assay system (Millipore), for which a cell monolayer was set in be-
tween a filter cell (apical) and a receiver plate (basolateral) at a den-
sity of 10000 cells per well. The culture medium was replaced
every 48 h, and the cells were kept in culture for 21 days. The trans
epithelial electrical resistance (TEER) of the monolayers was mea-
sured daily, before and after the experiment, by using an epithelial
volt–ohm meter (Millicell-ERS). Generally, TEER values greater than
CO at 378C. Mammalian cell lysis solution (50 mL) was added to all
2
wells, and the plate was shook for 5 min in an orbital shaker. Sub-
strate solution (50 mL) from the assay kit was added to all wells,
and the plate was shook for 5 min in an orbital shaker. The plate
was adapted in the dark for 10 min and the luminescence was
measured.
1
000” for a 21 day culture were considered optimal. Apical to ba-
Metabolism experiment in microsomes: Human liver microsomes
were purchased from BD Biosciences (Woburn, MA), NADPH was
purchased by Sigma–Aldrich. The experiment was performed fol-
lowing the protocol available on the website of BD Biosciences
solateral (Papp, A!B) and basolateral to apical (Papp, B!A) permea-
bility of the drug were measured at 120 min and at various drug
concentrations (1–100 mm). After 21 days of Caco-2 cell growth, the
medium was removed from the filter wells and from the receiver
plate, which were filled with fresh Hank’s balanced salt solution
(
www.bdbiosciences.com). Tested compounds (14 mL of 4 mm solu-
À1
tion in DMSO) were incubated with 20 mgmL human liver micro-
somal proteins (70 mL) in PBS (2.6 mL) at 378C for 5 min. NADPH
solution (20 mm) was freshly made in PBS and was added (140 mL)
to the tubes in final concentration of 1 mm to start the reaction.
The first sample (400 mL) was taken immediately, and the reaction
was stopped by the addition of acetonitrile+0.1% formic acid
(
HBSS) buffer (Invitrogen). This procedure was repeated twice, and
the plates were incubated at 378C for 30 min. After incubation
time, the HBSS buffer was removed and compounds were added
to the filter well at various concentrations (1–100 mm), whereas
fresh HBSS was added to the receiver plate. The plates were incu-
bated at 378C for 120 min. Afterward, samples were removed from
the apical (filter well) and basolateral (receiver plate) side of the
monolayer to measure the permeability. The apparent permeability
(
800 mL). Other samples were taken at time points of 15, 30, 45, 60,
9
1
0, and 120 min. Three control tubes were also incubated for
20 min: one without microsomes, one without NADPH, and one
(Papp) was calculated by using Equation (1):
without test compound. In these tubes, volume was replaced by
PBS. Sample tubes were vortexed and centrifuged for 6 min at
12000 rpm. Samples were injected into the UPLC–MS–MS (Waters
Xevo QTOF MS, column ACQUITY UPLC BEH C18 1.7 mm, eluent
VA
½drugŠacceptor
Papp ½nm sÀ1Š ¼ ð
Þ Â ð
Þ
ð1Þ
area  time
½drugŠ
initial
À1
H O/CH CN + 0.1% HCOOH, flow 0.6 mLmin , ESI positive mode).
2
3
in which V is the volume in milliliters in the acceptor well, area is
the surface area of the membrane (0.11 cm of the well), time is
The Metabolynx XS software (Waters) was used for the analysis of
metabolites.
A
2
the total transport time in seconds (7200 s), [drug]acceptor is the con-
centration of the drug measured by UV spectroscopy, and [dru-
À4
g]initial is the initial drug concentration (1”10 m) in the apical or
basolateral wells.
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
116
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
11
18
18
In vitro experiments of [ C]-5, [ F]-7, and [ F]-6
Keywords: glycoproteins · inhibitors · isotope labeling ·
positron emission tomography · radioligands
Separation of subcellular fractions for detecting fraction-specific
tracer uptake/binding: The following procedure was modified from
the original subcellular fractionation protocol available on the web-
site of Abcam (www.abcam.com). At the time of use: 10 mL (de-
pending on the number of samples) of the subcellular fraction
buffer were taken and 1 mm DTT (50 mL) and protease inhibitor
cocktail (50 mL) were added. After pre-incubation and incubation
with the above-mentioned modulators and/or ATP, the media con-
taining the radiotracer was removed from the cells. Three to four
samples of these radioactive media were transferred into clean
prelabeled Eppendorf tubes to use as an injected-dose (ID) correc-
tion. The remaining cells in the plate were harvested by trypsiniza-
tion and were collected into clean Eppendorf tubes containing
media (without serum) to neutralize the trypsin. This cell suspen-
sion was readily centrifuged at 6000”g for 2 min to obtain the cell
pellet. After removing all the supernatant, the cold subcellular frac-
tion buffer (500 mL) was added to each cell pellet. Using distinct
[
1] P. Kannan, C. John, S. S. Zoghbi, C. Haldldin, M. M. Gottesman, R. Gottes-
man, B. Innis, M. D. Hall, Clin. Pharmacol. Ther. 2009, 86, 368–377.
2] C. L. Graff, G. M. Pollack, Curr. Drug Metab. 2004, 5, 95–108.
[
[
3] J. R. Cirrito, R. Deane, A. M. Fagan, M. L. Spinner, M. Parsadanian, M. B.
Finn, H. Jiang, J. L. Prior, A. Sagare, K. R. Bales, S. M. Paul, B. V. Zlokovic,
D. Piwnica-Worms, D. M. Holtzman, J. Clin. Invest. 2005, 115, 3285–
3290.
[4] A. L. Bartels, Curr. Pharm. Des. 2011, 17, 2771–2777.
[5] A. H. Abuznait, A. Kaddoumi, Chem. Neurosci. 2012, 3, 820–831.
[6] A. Elali, S. Rivest, Front. Physiol. 2013, 13, 1–6.
[
7] C. Kuntner, J. P. Bankstahl, M. Bankstahl, J. Stanek, T. Wanek, G. Stund-
ner, R. Karch, R. Brauner, M. Meier, X. Ding, M. Müller, W. Lçscher, O.
Langer, Eur. J. Nucl. Med. Mol. 2010, 37, 942–953.
[
8] M. Bauer, M. Zeitlinger, R. Karch, P. Matzneller, J. Stanek, W. Jäger, M.
Bçhmdorfer, W. Wadsak, M. Mitterhauser, J. P. Bankstahl, W. Lçscher, M.
Koepp, C. Kuntner, M. Müller, O. Langer, Clin. Pharmacol. Ther. 2012, 91,
2
5G needles and 1 mL syringes to avoid cross contamination, the
cell pellets were homogenized 10 times for each sample. Until all
tubes were homogenized, the tubes were kept on ice. After ho-
mogenization, another set up of centrifugation at 500”g for 5 min
was performed. The supernatants (cytoplasmic extract) were then
immediately transferred into clean prelabeled tubes on ice and
kept until gamma counting. To the remaining pellet, more of the
cold subcellular fraction buffer (500 mL) was added, and it was vor-
texed for 5 s on the highest setting and kept on ice. After all tubes
were vortexed, another centrifugation set up of 3000”g for 5 min
was performed. The supernatants (membrane extract) were then
transferred into clean prelabeled tubes on ice. The remaining
pellet, which was called the “rest of the cells”, was also counting in
the gamma counter.
2
27–233.
[9] a) A. L. Bartels, O. L. de Klerk, R. Kortekaas, J. J. de Vries, K. L. Leenders,
Curr. Top. Med. Chem. 2010, 10, 1775–1784; b) W. C. Kreisl, J. Liow, S. N.
Kimura, N. Seneca, S. S. Zoghbi, C. L. Morse, P. Herscovitch, V. W. Pike,
R. B. Innis, J. Nucl. Med. 2010, 51, 559–566.
10] M. Lubberink, G. Luurtsema, B. N. van Berckel, R. Boellaard, R. Toornvliet,
A. D. Windhorst, E. J. Franssen, A. A. Lammertsma, J. Cereb. Blood Flow
Metab. 2007, 27, 424–433.
11] G. Luurtsema, C. F. Molthoff, R. C. Schuit, A. D. Windhorst, A. A. Lam-
mertsma, E. J. Franssen, Nucl. Med. Biol. 2005, 32, 87–93.
12] H. Nakanishi, A. Yonezawa, K. Matsubara, I. Yano, Eur. J. Pharmacol.
2013, 710, 20–28.
[13] T. Wanek, S. Mairinger, O. Langer, J. Labelled Compd. Radiopharm. 2013,
56, 68–77.
[
[
[
[
14] G. Luurtsema, R. C. Schuit, R. P. Klok, J. Verbeek, J. E. Leysen, A. A. Lam-
Transport assay using the radiolabeled compounds: Complete RPMI
medium including l-glutamine and FCS (fetal calf serum) were
used throughout. Cells were seeded on microporous poly-
carbonate membrane filters (8.0 mm pore size, 6.5 mm diameter,
mertsma, A. D. Windhorst, Nucl. Med. Biol. 2009, 36, 643–649.
[15] P. Kannan, S. Telu, S. Shukla, S. V. Ambudkar, V. W. Pike, C. Halldin, M. M.
Gottesman, R. B. Innis, M. D. Hall, ACS Chem. Neurosci. 2011, 2, 82–89.
[16] N. A. Colabufo, A. van Waarde, Curr. Top. Med. Chem. 2010, 10, 1701–
1702.
Transwell 3422, Costar Corp., Cambridge, MD) at a density of 3”
4
[17] S. Mairinger, T. Wanek, C. Kuntner, Y. Doenmez, S. Strommer, J. Stanek,
E. Capparelli, P. Chiba, M. Müller, N. A. Colabufo, O. Langer, Nucl. Med.
Biol. 2012, 39, 1219–1225.
18] A. van Waarde, N. K. Ramakrishnan, A. A. Rybczynska, P. H. Elsinga, F. Be-
rardi, J. R. de Jong, C. Kwizera, R. Perrone, M. Cantore, J. W. Sijbesma,
R. A. Dierckx, N. A. Colabufo, J. Med. Chem. 2009, 52, 4524–4532.
[19] N. Tournier, H. Valette, M. A. Peyronneau, W. Saba, S. Goutal, B. Kuhnast,
F. DollØ, J. M. Scherrmann, S. Cisternino, M. Bottlaender, J. Nucl. Med.
2011, 52, 415–423.
[20] S. D. Bruyne, L. Wyffels, L. Moerman, J. Sambre, N. A. Colabufo, F. Berar-
di, R. Perrone, F. D. Vos, Bioorg. Med. Chem. 2010, 18, 6489–6495.
1
0 cells per well. The cells were grown for 3 days (72 h) in com-
plete medium with one medium replacement. At 1–2 h before the
start of the experiment, medium at both the apical and the basal
side of the monolayer was replaced with a total of 700 mL (100 mL
in the apical and 600 mL in the basal side) of complete medium.
[
[30]
The experiment was started by adding P-gp modulator 3, 1, or
3 (stock solution of 10 mm) in the apical side of the cell layer for
1
3
6
0 min and was followed by the addition of the radiotracer (15 mL,
.6 kBqmL ), which was incubated for more 30 min at 378C under
À1
an atmosphere of 5% CO with regulated humidity. No P-gp mod-
2
[
[
[
21] N. A. Colabufo, F. Berardi, M. G. Perrone, E. Capparelli, M. Cantore, C. In-
glese, R. Perrone, Curr. Top. Med. Chem. 2010, 10, 1703–1714.
22] R. B. Wang, C. L. Kuo, L. L. Lien, E. J. Lien, J. Clin. Pharm. Ther. 2003, 28,
ulator was added in the controls (only the radiolabeled com-
pound). After the incubating period, all media from each compart-
ment was removed and transferred to clean Eppendorf tubes for
subsequent radioactivity measures in the gamma counter. The per-
centage of the injected dose (%ID) was then calculated.
203–228.
23] J. W. Polli, S. A. Wring, J. E. Humphreys, L. Huang, J. B. Morgan, L. O.
Webster, C. S. Serabjit-Singh, J. Pharmacol. Exp. Ther. 2001, 299, 620–
638.
[
[
24] N. A. Colabufo, F. Berardi, M. G. Perrone, M. Cantore, M. Contino, C. In-
glese, M. Niso, R. Perrone, ChemMedChem 2009, 4, 188–195.
25] A. Calcabrini, J. M. García-Martínez, L. Gonzµlez, M. J. Tendero, M. T.
OrtuÇo, P. Crateri, A. Lopez-Rivas, G. Arancia, P. Gonzµlez-PorquØ, J.
Martín-PØrez, Carcinogenesis 2006, 27, 1699–1712.
Acknowledgements
[
26] N. A. Colabufo, F. Berardi, M. Cantore, M. G. Perrone, M. Contino, C. In-
glese, M. Niso, R. Perrone, A. Azzariti, G. M. Simone, L. Porcelli, A. Paradi-
so, Bioorg. Med. Chem. 2008, 16, 362–373.
The UPLC–MS-QTOF facility is supported financially by The Neth-
erlands Organisation for Health Research and Development
[
[
27] D. M. Jewett, Appl. Radiat. Isot. 1992, 43, 1383–1385.
28] T. D. Wegman, B. Maas, P. H. Elsinga, W. Vaalburg, Appl. Radiat. Isot.
2002, 57, 505–507.
(
ZonMw) project number: 91111007. We thank Bram Maas for his
assistance with the radiochemistry.
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
117
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
[
29] a) B. Feng, J. B. Mills, R. E. Davidson, R. J. Mireles, J. S. Janiszewski, M. D.
Troutman, S. M. De Morais, Drug Metab. Dispos. 2008, 36, 268–275;
b) O. A. Fahmi, T. S. Mauer, M. Kish, E. Cardenas, S. Boldt, D. Nettleton,
Drug Metab. Dispos. 2008, 36, 1698–1708.
30] J. Rautio, J. E. Humphreys, L. O. Webster, A. Balakrishnan, J. P. Keogh,
J. R. Kunta, C. J. Serabjit-Singh, J. W. Polli, Drug Metab. Dispos. 2006, 34,
[31] A. L. Kangas, L. Groenroos, M. Nieminen, Med. Biol. 1984, 62, 338–343.
Received: September 14, 2015
Revised: October 19, 2015
7
86–792.
Published online on November 13, 2015
ChemMedChem 2016, 11, 108 – 118
www.chemmedchem.org
118
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim