Synthesis and Preclinical Evaluation of Three Novel Fluorine-18 Labeled Radiopharmaceuticals for P-Glycoprotein PET Imaging at the Blood-Brain Barrier

Article abstract of DOI:10.1021/mp5008103

P-Glycoprotein (P-gp), along with other transporter proteins at the blood-brain barrier (BBB), limits the entry of many pharmaceuticals into the brain. Altered P-gp function has been found in several neurological diseases. To study the P-gp function, many positron emission tomography (PET) radiopharmaceuticals have been developed. Most P-gp radiopharmaceuticals are labeled with carbon-11, while labeling with fluorine-18 would increase their applicability due to longer half-life. Here we present the synthesis and in vivo evaluation of three novel fluorine-18 labeled radiopharmaceuticals: 4-((6,7-dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)methyl)-2-(4-fluorophenyl)oxazole (1a), 2-biphenyl-4-yl-2-fluoroethoxy-6,7-dimethoxy-1,2,3,4-tetrahydro-isoquinoline (2), and 5-(1-(2-fluoroethoxy))-[3-(6,7-dimethoxy-3,4-dihydro-1H-isoquinolin-2-yl)-propyl]-5,6,7,8-tetrahydronaphthalen (3). Compounds were characterized as P-gp substrates in vitro, and Mdr1a/b^(-/-)Bcrp1^(-/-) and wild-type mice were used to assess the substrate potential in vivo. Comparison was made to (R)-[¹¹C]verapamil, which is currently the most frequently used P-gp substrate. Compound [¹⁸F]3 was performing the best out of the new radiopharmaceuticals; it had 2-fold higher brain uptake in the Mdr1a/b^(-/-)Bcrp1^(-/-) mice compared to wild-type and was metabolically quite stable. In the plasma, 69% of the parent compound was intact after 45 min and 96% in the brain. Selectivity of [¹⁸F]3 to P-gp was tested by comparing the uptake in Mdr1a/b^(-/-) mice to uptake in Mdr1a/b^(-/-)Bcrp1^(-/-) mice, which was statistically not significantly different. Hence, [¹⁸F]3 was found to be selective for P-gp and is a promising new radiopharmaceutical for P-gp PET imaging at the BBB. (Chemical Equation).

Full text of DOI:10.1021/mp5008103

Article
pubs.acs.org/molecularpharmaceutics
Synthesis and Preclinical Evaluation of Three Novel Fluorine-18
Labeled Radiopharmaceuticals for P‑Glycoprotein PET Imaging at
the Blood−Brain Barrier
Heli Savolainen,† Mariangela Cantore,‡,§ Nicola A. Colabufo,‡,§ Philip H. Elsinga,† Albert D. Windhorst,∥
and Gert Luurtsema*,†
†Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen,
Hanzeplein 1, 9713 GZ Groningen, Netherlands
‡Dipartimento di Farmacia-Scienze del Farmaco, Universita
§Biofordrug slr, via Orabona 4, 70125 Bari, Italy
̀
degli Studi di Bari, via Orabona 4, 70125 Bari, Italy
∥Department of Radiology and Nuclear Medicine, VU University Medical Center Amsterdam, De Boelelaan 1085 C, 1081 HV
Amsterdam, Netherlands
ABSTRACT: P-Glycoprotein (P-gp), along with other trans-
porter proteins at the blood−brain barrier (BBB), limits the
entry of many pharmaceuticals into the brain. Altered P-gp
function has been found in several neurological diseases. To
study the P-gp function, many positron emission tomography
(PET) radiopharmaceuticals have been developed. Most P-gp
radiopharmaceuticals are labeled with carbon-11, while
labeling with fluorine-18 would increase their applicability
due to longer half-life. Here we present the synthesis and in
vivo evaluation of three novel fluorine-18 labeled radiophar-
maceuticals: 4-((6,7-dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)methyl)-2-(4-fluorophenyl)oxazole (1a), 2-biphenyl-4-yl-2-
fluoroethoxy-6,7-dimethoxy-1,2,3,4-tetrahydro-isoquinoline (2), and 5-(1-(2-fluoroethoxy))-[3-(6,7-dimethoxy-3,4-dihydro-1H-
isoquinolin-2-yl)-propyl]-5,6,7,8-tetrahydronaphthalen (3). Compounds were characterized as P-gp substrates in vitro, and
Mdr1a/b(−/−)Bcrp1(−/−) and wild-type mice were used to assess the substrate potential in vivo. Comparison was made to (R)-
[11C]verapamil, which is currently the most frequently used P-gp substrate. Compound [18F]3 was performing the best out of the
new radiopharmaceuticals; it had 2-fold higher brain uptake in the Mdr1a/b(−/−)Bcrp1(−/−) mice compared to wild-type and was
metabolically quite stable. In the plasma, 69% of the parent compound was intact after 45 min and 96% in the brain. Selectivity of
[18F]3 to P-gp was tested by comparing the uptake in Mdr1a/b(−/−) mice to uptake in Mdr1a/b(−/−)Bcrp1(−/−) mice, which was
statistically not significantly different. Hence, [18F]3 was found to be selective for P-gp and is a promising new
radiopharmaceutical for P-gp PET imaging at the BBB.
KEYWORDS: blood−brain barrier, P-glycoprotein, BCRP, PET, 18F, radiopharmaceutical
(PD),7 while it has been hypothesized that P-gp function is
1. INTRODUCTION
increased in epilepsy.8
P-Glycoprotein (P-gp) is an efflux transporter protein on the
Positron emission tomography (PET) is a molecular imaging
luminal membrane of the brain endothelial cells.1 It belongs to
technique that is able to measure tissue concentrations of
a family of transporters that has a similar function at the
blood−brain barrier (BBB), where P-gp and the Breast Cancer
compounds labeled with positron emitting isotopes as a
function of time in a noninvasive manner. To study P-gp
Resistance Protein (BCRP) are the most important in the
function with PET, radiolabeled substrates as well as radio-
human brain.2 This family of transporters is also known as the
ATP-binding cassette (ABC) transporters, as their function is
ATP dependent.3 They recognize small hydrophobic exoge-
nous compounds that have diffused through the BBB and efflux
them back into the blood. Consequently, P-gp not only
protects the brain from harmful compounds, but can also
reduce the uptake of drugs that need to act in the brain.4
Moreover, P-gp function is altered in some neurological
diseases. For instance, decreased P-gp function has been
found in Alzheimer’s disease (AD)5,6 and Parkinson’s disease
labeled inhibitors could be applied.9 The most widely used
substrate radiopharmaceutical for P-gp imaging is [11C]-
verapamil. It has been used either as racemic mixture or as
(R)-enantiomer10 and has been utilized in clinical research.11,12
The P-gp inhibitors [11C]tariquidar and [11C]elacridar were
Received: December 2, 2014
Revised: May 12, 2015
Accepted: May 22, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
Chart 1. Structures of [18F]1a, [18F]2, [18F]3, [18F]Fluoroethyl Tariquidar, [18F]Fluoroethyl Elacridar and 1-
[18F]Fluoroelacridar
aimed to bind P-gp with high selectivity. However, they were
proved to behave as substrates at tracer level and are not
specific for P-gp, but are also transported by BCRP.13,14
Most of the labeled compounds that are currently in use are
labeled with carbon-11, while labeling with fluorine-18 would
lead to a radiopharmaceutical that can be transported to other
imaging centers without on-site cyclotron due to the longer
half-life of fluorine-18 of 110 min vs 20 min for carbon-11. In
addition, the longer half-life would enable prolonged imaging
times and also more subjects could be dosed from one
radiopharmaceutical production. Half-life of fluorine-18 can be
a disadvantage only if repeated scanning for the same subject
during one day is desired.
There are only a few examples of fluorine-18 labeled P-gp
radiopharmaceuticals that have been applied in CNS disorders.
Tariquidar and elacridar analogues have been labeled with
fluorine-18 by a fluoroalkylation method, demonstrating similar
in vivo behavior as their carbon-11 labeled counterparts.15 An
elacridar analogue has been labeled also via a nucleophilic
aromatic fluorination, but the utilization of 1-[18F]-
fluoroelacridar is limited due to the low radiochemical yield
and in vivo defluorination.16
2. EXPERIMENTAL SECTION
2.1. Chemicals and Cell Lines. All reagents and solvents
were obtained from commercial suppliers (Merck (New Jersey,
USA), Sigma-Aldrich (St. Louis, USA), Rathburn (Walkerburn,
UK), and B. Braun (Melsungen, Germany)) and were used
without further purification. (R)-Norverapamil was purchased
from ABX advanced biochemical compounds (Radeberg,
Germany). Human liver microsomes were obtained from BD
Gentest (20 mg/mL) (San Jose, USA). Cell culture reagents
were purchased from Celbio s.r.l. (Pero Mi, Italy). CulturePlate
96/wells plates were bought from PerkinElmer Life Science
(Waltham, USA). MDCK-MDR1, MDCK-BCRP, and MDCK-
MRP1 cell lines were a gift from Professor P. Borst, NKI-AVL
Institute, Amsterdam, Netherlands. Caco-2 cells were a gift
from Dr. Aldo Cavallini and Dr. Caterina Messa from the
Laboratory of Biochemistry, National Institute for Digestive
Diseases, ‘S. de Bellis’, Bari, Italy.
2.2. Organic Synthesis. 2.2.1. General Methods. Column
chromatography was performed with Merck silica gel 60 Å
(63−200 μm) as the stationary phase. Recording of mass
spectra was done on an HP 6890-5973MSD gas chromato-
graph/mass spectrometer (Santa Clara, USA); only significant
m/z peaks, with their percentage of relative intensity in
Clearly, there is an unmet need to develop 18F-radiolabeled
compounds to study P-gp function with PET at the BBB. We
report three novel compounds: 4-((6,7-dimethoxy-3,4-dihy-
droisoquinolin-2(1H)-yl)methyl)-2-(4-fluorophenyl)oxazole
(1a), 2-biphenyl-4-yl-2-fluoroethoxy-6,7-dimethoxy-1,2,3,4-tet-
rahydro-isoquinoline (2), and 5-(1-(2-fluoroethoxy))-[3-(6,7-
dimethoxy-3,4-dihydro-1H-isoquinolin-2-yl)-propyl]-5,6,7,8-
tetrahydronaphthalen (3), which were identified as P-gp
substrates in in vitro experiments, as presented in this Article.
All of the compounds contain the same basic 6,7-dimethox-
ytetrahydroisoquinoline moiety as tariquidar or elacridar also
have.17,18 All three compounds were labeled with fluorine-18
(Chart 1), and their brain uptake was investigated with PET
imaging in FVB mice as well as in the Mdr1a/b(−/−)Bcrp1(−/−)
knockout mice, and results were compared to those of (R)-
[11C]verapamil. In addition, [18F]3 was investigated in Mdr1a/
b(−/−) mice to see if the compound was selective for P-gp.
1
parentheses, are reported. H NMR spectra were recorded in
CDCl3 at 300 MHz with Varian Mercury-VX spectrometer
(Palo Alto, USA) or on an Avance 500 MHz spectrometer
(Bruker, Billerica, USA). All chemical shift values are reported
in ppm (δ) relative to the solvent peak (7.27 for CHCl3).
Melting points were determined in open capillary on a
Gallenkamp electrothermal apparatus (Leicestershire, UK).
ESI-MS analyses were performed on an Agilent 1100 LC/
MSD trap system VL (Santa Clara, USA). Elemental analyses
(C, H, N) were performed on Eurovector Euro EA 3000
analyzer (Milan, Italy); the analytical results were within 0.4%
of the theoretical values for the given formula.
2.2.2. Synthesis of 4-Fluorobenzamide (5a). 4-Fluoroben-
zoic acid (4a, 2.0 g, 14 mmol) was refluxed with SOCl2 (2.0
mL, 28 mmol) in the presence of Et3N (1.0 mL, 7.2 mmol) for
1 h. After evaporation of SOCl2, the resulting acyl chloride was
B
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
added to a 45 mL mixture of NH4OH (8 M), H2O, and CH2Cl2
(1:1:1 v/v/v). This mixture was stirred at room temperature for
4 h. The organic layer was separated from the aqueous layer
and washed with 2 M NaOH (3 × 10 mL). The organic
solution was dried over Na2SO4 and evaporated under reduced
pressure. The residue was purified by silica gel column
chromatography (CH2Cl2/MeOH 95:5 v/v). The yield of
brown solid 5a was 590 mg (30%). GC−MS m/z: 139 (M+,
67), 123 (100), 95 (72).
reduced pressure. The residue was purified on silica gel column
chromatography (CHCl3/MeOH 19:1 v/v) and recrystallized
from MeOH/Et2O to obtain 190 mg of 2 (45%). (C,N,H)
C26H28FNO3HCl. ESI+/MS m/z: 444 [M + Na]+. ESI+/MS/
MS m/z: 441 (100). 1H NMR (300 MHz, CDCl3) δ: 7.42 (dd,
4H, J = 7.2 Hz, J = 8.0 Hz phenyl-H2′,3′,5′,6′), 7.12 (d, 2H, J = 8
Hz, phenyl-H2,6), 7.01 (d, 2H, J = 7.5 Hz, phenyl-H3,5), 6.60 (s,
1H, isoquinoline-H5), 6.45 (s, 1H, isoquinoline-H8), 4.83 (t,
1H, J = 3 Hz, FCH2CH2), 4.60 (t, 1H, J = 3 Hz, FCH2CH2),
4.14 (t, 1H, J = 3 Hz, FCH2CH2), 4.10 (t, 1H, J = 3 Hz,
FCH2CH2), 3.83 (s, 3H, CH3), 3.80 (s, 3H, CH3), 3.71 (s, 2H,
NCH2), 3.57 (s, 2H, NCH2) 2.83−2.76 (m, 4H, NCH2CH2).
2.2.6. Synthesis of 5-(1-(2-Fluoroethoxy))-[3-(6,7-dime-
thoxy-3,4-dihydro-1H-isoquinolin-2-yl)-propyl]-5,6,7,8-tetra-
hydronaphthalen (3). A suspension of NaH (48 mg, 2.0
mmol) in dry DMF (3.0 mL) was stirred at room temperature
for 10 min. A solution of phenol precursor 14 (560 mg, 2.0
mmol) in DMF (1.0 mL) was added, and the solution was
stirred for 1 h. A solution of 2-fluoroethyl tosylate was added
(1.1 g, 4.0 mmol) in DMF (1.0 mL), and the reaction mixture
was stirred for 4 h. Water was added until effervescence ceased.
The solvent was evaporated, and the residue was partitioned
between H2O (20 mL) and CHCl3 (20 mL). The organic phase
was separated, the aqueous phase was extracted with CHCl3 (3
× 50 mL), and the collected organic fractions were dried over
Na2SO4 and evaporated under reduced pressure. The residue
was purified on silica gel column chromatography (CHCl3/
MeOH 19:1 v/v) and recrystallized from MeOH/Et2O to yield
320 mg of 3 (38%). (C,N,H) C26H35FNO3HCl. ESI+/MS m/z:
2.2.3. Synthesis of 4-(Chloromethyl)-2-(4-fluorophenyl)-
oxazole (6a). Amide 5a (590 mg, 4.2 mmol) was reacted with
1,3-dichloroacetone (1.0 g, 8.4 mmol) at 200 °C for 5 h. After
cooling to room temperature, water (10 mL) and CHCl3 (10
mL) were added. The aqueous phase was extracted with CHCl3
(3 × 50 mL), and the combined organic layers were dried over
Na2SO4 and evaporated under reduced pressure. The residue
was purified on silica gel column chromatography (hexane/
EtOAc 8:2 v/v) yielding 270 mg (30%) of 6a as a brown solid.
1
GC−MS m/z: 213 (M+ + 2, 19), 211 (M+, 57), 176 (78). H
NMR (300 MHz, CDCl3) δ: 8.06 (d, 1H, J = 8.0 Hz, phenyl-
H2), 8.03 (d, 1H, J = 8.1 Hz, phenyl-H3), 7.60 (s, 1H, oxazole-
H5), 7.18 (d, 1H, J = 8.3 Hz, phenyl-H6), 7.16 (d, 1H, J = 8.0
Hz, phenyl-H5), 4.32 (s, 2H, CH2).
2.2.4. Synthesis of 4-((6,7-Dimethoxy-3,4-dihydroisoqui-
nolin-2(1H)-yl)methyl)-2-(4-fluorophenyl)oxazole (1a). Com-
pound 6a (260 mg, 1.0 mmol) was alkylated with 6,7-
dimethoxy-1,2,3,4-tetrahydroisoquinoline (460 mg, 2.4 mmol)
in DMF (20 mL) using Na2CO3 as a base (250 mg, 2.4 mmol).
The mixture was refluxed overnight. DMF was evaporated
under reduced pressure, and the residue was partitioned
between H2O (20 mL) and CHCl3 (20 mL). The organic phase
was separated, the aqueous phase was extracted with CHCl3 (3
× 50 mL), and the collected organic fractions were dried over
Na2SO4 and evaporated under reduced pressure. The residue
was purified on silica gel column chromatography (CH2Cl2/
EtOAc 1:1 v/v) to obtain 270 mg of 1a as a brown oil in 74%
yield. Hydrochloride salt was dissolved in methanol while
heating, Et2O was added, and the solution was cooled down to
recrystallize the product. Mp: 232−235 °C. ESI+/MS m/z: 369
[M + H]+, 391 [M + Na]+. ESI+/MS/MS m/z: 352 (16), 230
1
450 [M + Na]+. ESI+/MS/MS m/z: 420 (47), 388 (100). H
NMR (300 MHz, CDCl3) δ: 7.08 (dd, 1H, J = 7.8 Hz, J = 6 Hz,
phenyl-H7), 6.82 (d, 1H, J = 7.7 Hz, phenyl-H6), 6.63 (d, 1H, J
= 7 Hz, phenyl-H8), 6.58 (s, 1H, isoquinoline-H5), 6.50 (s, 1H,
isoquinoline-H8), 4.83 (t, 1H, J = 3 Hz, FCH2CH2), 4.66 (t,
1H, J = 3 Hz, FCH2CH2), 4.24 (t, 1H, J = 3 Hz, FCH2CH2),
4.14 (t, 1H, J = 3 Hz, FCH2CH2), 3.84 (s, 3H, CH3), 3.83 (s,
3H, CH3), 3.55 (s, 2H, NCH2), 2.60−2.84 (m, 9H,
tetrahydronaphthalen-CH2CH2CH2CH, isoquinoline-
NCH2CH2CH2, NCH2CH2), 1.63−1.85 (m, 8H, tetrahydro-
naphthalen-CH2CH2CH2CH, isoquinoline-NCH2CH2CH2).
2.2.7. Synthesis of 2-Fluoroethyl Tosylate. Synthesis of 2-
fluoroethyl tosylate was performed by making a solution of 2-
fluoroethanol (1.3 mL, 2.0 mmol) and p-toluenesulfonyl
chloride (170 mg, 1.1 mmol) in 5 M NaOH (5.0 mL, 1.6
mmol), which was stirred at room temperature for 24 h. The
reaction mixture was diluted with CH2Cl2, and the organic
phase was washed with 10% NaOH. The organic layer was
dried (Na2SO4) and evaporated under reduced pressure. The
crude product was purified by chromatography on a silica gel
column with CH2Cl2 to give 220 mg (92%) of colorless oil that
was stored in a freezer. ESI+/MS m/z: 241 [M + Na]+. ESI+/
MS/MS m/z: 241 (74), 97 (100). 1H NMR (300 MHz,
CDCl3) δ: 7.80 (d, 2H, J = 8 Hz, phenyl-H2,6), 7.34 (d, 2H, J =
8 Hz, phenyl-H3,5), 4.49 (t, 1H, J = 4 Hz, CH2), 4.64 (t, 1H, J =
4 Hz, CH2), 4.30 (t, 1H, J = 4 Hz, CH2), 4.21 (t, 1H, J = 4 Hz,
CH2), 2.45 (s, 3H, CH3).
1
(50), 176 (98), 121 (100). H NMR (300 MHz, CDCl3) δ:
8.05 (d, 1H, J = 8.0 Hz, phenyl-H2) 8.03 (d, 1H, J = 8.1 Hz,
phenyl-H3), 7.60 (s, 1H, oxazole-H5), 7.16 (d, 1H, J = 8.0 Hz,
phenyl-H5), 7.15 (d, 1H, J = 8.2 Hz, phenyl-H6), 6.58 (s, 1H,
isoquinoline-H5), 6.50 (s, 1H, isoquinoline-H8), 3.88 (s, 3H,
CH3), 3.82 (s, 3H, CH3), 3.75−3.79 (m, 4H, NCH2CH2), 2.94
(s, 4H, CH2NCH2). Purity of the final compound was
established by combustion analysis of the corresponding
hydrochloride salt, confirming a purity ≥95%. (C,N,H)
C21H21FN2O3HCl(H2O)0.5.
2.2.5. Synthesis of 2-Biphenyl-4-yl-2-fluoroethoxy-6,7-
dimethoxy-1,2,3,4-tetrahydro-isoquinoline (2). A suspension
of NaH (24 mg, 1.0 mmol) in dry DMF (3.0 mL) was stirred at
room temperature for 10 min. A solution of phenol precursor 9
(1.0 g, 1.0 mmol) in DMF (1.0 mL) was added, and the
solution was stirred for 1 h. A solution of 2-fluoroethyl tosylate
was added (860 mg, 2.0 mmol) in DMF (1.0 mL), and the
reaction mixture was stirred for 4 h. Water was added until
effervescence ceased. The solvent was evaporated, and the
residue was partitioned between H2O (20 mL) and CHCl3 (20
mL). The organic phase was separated, the aqueous phase was
extracted with CHCl3 (3 × 50 mL), and the collected organic
fractions were dried over Na2SO4 and evaporated under
2.2.8. Synthesis of 2-Bromoethyl Tosylate (15). A volume
of 3.0 mL of DCM was added to toluenesulfonyl chloride (950
mg, 5.0 mmol) in a round-bottom flask, and the mixture was
brought to 0 °C. Triethylamine (680 μL, 5.0 mmol) was added,
and 2-bromoethanol (280 μL, 4.0 mmol) was added dropwise.
The mixture was stirred for 1 h at 0 °C. It was brought to room
temperature, washed with water (10 mL), brine (10 mL), and
C
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
dried over Na2SO4. Crude product was concentrated under
reduced pressure into brown liquid. Product was purified by
silica gel column chromatography (hexane/EtOAc 9:1 v/v).
Product was isolated as a colorless liquid (550 mg, yield 49%).
1H NMR (500 MHz, CDCl3) δ: 7.84 (d, 2H, J = 5 Hz, phenyl-
H2,6), 7.39 (d, 2H, J = 10 Hz, phenyl-H3,5), 4.31 (t, 2H, J = 5
Hz, CH2), 3.50 (t, 2H, J = 7 Hz, CH2), 2.48 (s, 3H, CH3).
2.3. Radiochemistry. 2.3.1. Production and Workup of
Fluorine-18. [18F]Fluoride was produced by irradiation of
[18O]water with the Scanditronix MC-17 cyclotron (Uppsala,
Sweden) via the 18O(p,n)18F nuclear reaction. The [18O]water
containing the [18F]fluoride was passed over a pretreated Sep-
Pak light Accell Plus QMA (Waters, Milford, USA) (pretreated
with 5 mL of 1.4% Na2CO3 solution (B. Braun, Melsungen,
Germany), washed with water until pH was neutral, and dried
using argon). The [18F]fluoride was eluted from the cartridge
with solution of K2CO3 (5.0 mg, 36 mmol in production of
[18F]1a, and 3.0 mg, 22 mmol in production of [18F]2 and
[18F]3 in 1.0 mL of water) and collected into a vial containing
Kryptofix[2.2.2] (15 mg, 40.0 mmol). A volume of 1 mL of dry
acetonitrile was added, and solvents were evaporated to dryness
at 130 °C with a helium flow (200 mL/min). Next, 0.5 mL of
acetonitrile was added, and the mixture was dried again under
the same conditions. This procedure was repeated another two
times.
3 10 min). The product was collected into a bottle containing
80 mL of sterile water. After mixing with helium, the solution
was passed over an Oasis HLB 1 cm3 (30 mg) extraction
cartridge to trap the product, after which the cartridge was
washed twice with 8 mL of water. Finally the product was
eluted from the cartridge with 1 mL of ethanol and passed
through a Millipore Millex LG filter (0.2 μm). The filtrate was
diluted with 4 mL of 0.9% NaCl. Analysis of the product was
performed by HPLC using an Alltech Alltima C18 5 μm 4.6 ×
250 mm as a column (Thermo Fisher Scientific, Waltham,
USA) using the same eluent as in prep. HPLC for [18F]2 and
MeCN/water 1:1 + 0.1% TFA (v/v) for [18F]3, both at a flow
of 1 mL/min.
2.3.4. Radiosynthesis of (R)-[11C]Verapamil. The synthesis
of (R)-[11C]verapamil was performed as described previously
with some modifications.19 Briefly, [11C]CH4, produced
directly in the target using 14N(p,α)11C nuclear reaction, was
trapped in the liquid nitrogen. [11C]Methane was first
converted into [11C]methyl iodide and further into [11C]-
methyl triflate by passage through a silver triflate/α-alumina
column (Sigma-Aldrich, St. Louis, USA).19 [11C]Methyl triflate
was bubbled into (R)-norverapamil (0.5 mg, 1.0 mmol)
solution in 0.5 mL of acetonitrile. The reaction mixture was
heated for 5 min at 120 °C. Water (0.5 mL) and MeCN (0.5
mL) were added, and the solution was injected into preparative
HPLC system (column Symmetryshield RPS 5 μm 7.8 × 300
mm) and purified with 25 mM NaH2PO4 pH 7.0/MeCN/
MeOH (41:37:22 v/v/v) as mobile phase (flow rate 5 mL/min,
UV detection at 210 nm, Rt((R)-[11C]verapamil) = 9 min).
The eluted fraction containing [11C]verapamil was collected
into 80 mL of sterile water. After mixing with helium, organic
solvents were removed by passing the mixture through an Oasis
HLB 1 cm3 (30 mg) extraction cartridge, following rinsing of
the cartridge (twice) with 8 mL of saline solution (0.9% NaCl).
(R)-[11C]Verapamil was eluted by passing 0.8 mL of EtOH
through the cartridge and Millipore Millex LG filter (0.2 μm)
and formulated with 4 mL of saline solution. Analysis of the
product was performed by HPLC using an Alltima C18 5 μm
4.6 × 250 mm column and 0.1 M NaH2PO4 (pH 3)/MeCN
(65:35 v/v) as eluent (flow rate 1.5 mL/min, UV detection at
210 nm).
2.3.2. Radiosynthesis of [18F]1a. The precursor 1b (2.0 mg,
5.0 mmol) was dissolved in 0.5 mL of DMF and was added to
the dried K[18F]F/Kryptofix complex. Mixture was reacted for
30 min at 160 °C and cooled for 2 min after which 0.5 mL of
HPLC eluent (0.05 M NaOAc/MeOH/THF 55:27:18 v/v/v)
was added. The total mixture was filtered through an Acrodisc
syringe filter (0.4 μm, Pall Life Sciences, Port Washington,
USA) and subjected to HPLC purification utilizing a
Symmetryshield RPS 5 μm 7.8 × 300 mm column (Waters,
Milford, USA) at a flow of 2 mL/min (UV detection at 254 nm,
Rt([18F]1a) = 15 min). The product was collected into a bottle
containing 80 mL of sterile water. After mixing with helium, the
solution was passed over an Oasis HLB 1 cm3 (30 mg)
extraction cartridge (Waters, Milford, USA) to trap [18F]1a,
after which the cartridge was washed twice with 8 mL of water.
Finally the product was eluted from the cartridge with 1 mL of
ethanol and passed through a Millipore Millex LG filter (0.2
μm, Billerica, USA). The filtrate was diluted with 4 mL of 0.9%
NaCl. Analysis of the product was performed by HPLC using
an Xterra C18 5 μm 4.6 × 250 mm column (Waters, Milford,
USA) and 0.05 M NaH2PO4/MeOH/THF 55:27:18 (v/v/v) as
eluent at a flow of 1 mL/min.
2.4. Distribution Coefficient Log D. n-Octanol (0.5 mL)
and phosphate buffered saline (PBS, pH = 7.2, 0.5 mL) were
pipetted in 1:1 ratio into Eppendorf tubes. The radio-
pharmaceutical solution (∼1 MBq, 100 μL) was added, and
tubes were vortexed for 1 min and centrifuged for 5 min at
6000 rpm. Samples (100 μL) of octanol and PBS layer were
counted on a γ-counter (LKBG-Compugamma CS 1282,
Wallac, Waltham, USA). The distribution coefficient log D
was calculated as log(Aoctanol/APBS). Measurements were done
in triplicate.
2.3.3. Radiosynthesis of [18F]2 and [18F]3. In the synthesis
of [18F]2 and [18F]3, 15 μL (21 mg, 75 μmol) of 2-bromoethyl
tosylate (15) in 1.0 mL of 1,2-dichlorobenzene (DCB) was
added to the dried fluoride complex. Distillation of the formed
[18F]bromoethyl fluoride (16) at 90° was started immediately
with a helium gas flow (40 mL/min) to the second vial in room
temperature containing 2.0 mg of precursor (5.0 mmol 9 or 14)
and 3.0 mg (75 mmol) of 60% dispersion of NaH in mineral oil
in 0.5 mL of DMF. After 15 min of distillation, vial 2 was
reacted for 5 min at 80 °C in the synthesis of [18F]3 and 10 min
in the synthesis of [18F]2. After completion of the reaction, 0.5
mL of HPLC eluent (0.1 M NaOAc/MeCN 5.5:4.5 v/v) was
added. The mixture was filtered through an Acrodisc syringe
filter (0.4 μm) and subjected to HPLC purification utilizing a
Symmetryshield RPS 5 μm 7.8 × 300 mm column at a flow of 3
mL/min (UV detection at 254 nm, Rt of both [18F]2 and [18F]
2.5. Animals. All animal studies were in compliance with
the local ethical guidelines, and protocols were approved by the
Institutional Animal Care and Use Committee of the University
of Groningen. Male FVB wild type mice (28 1.8 g), Mdr1a/
b
(−/−)Bcrp1(−/−) constitutive knockout mice (29 1.1 g), and
Mdr1a/b(−/−) constitutive knockout mice (28
2.3 g)
developed from the FVB line were purchased from Taconic
(Hudson, USA). After arrival, animals were acclimatized at least
7 days in the Central Animal Facility of the University Medical
Center Groningen. Mice had access to food and water ad
libitum and were kept under a 12 h light−dark cycle. During
D
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
a
Scheme 1. Synthesis of 1a,b
a
Reagents and conditions: (i) SOCl2, Et3N, NH4OH, CH2Cl2; (ii) 1,3-dichloroacetone; (iii) Na2CO3, DMF.
experiments, mice were anesthetized with 2% isoflurane in
medical air and warmed with a heating pad.
2.6. PET Procedure. Mice were injected with a fluorine-18
metabolites. Found mass peaks were integrated, and the area
was converted to percentage considering the parent area at T =
0 min as 100%. Percentage of the parent compound and found
metabolites were expressed as a function of time. Results from
2−3 experiments were fitted using second order polynomial
fitting in GraphPad Prism 5 software (La Jolla, USA).
radiopharmaceuticals (5.3 1.9 MBq, 0.1−0.2 mL, 64 22 ng
of 1a, 7.0
7.0 ng of 2, 0.3
0.01 ng of 3) or (R)-
180 ng of
[11C]verapamil (7.3
3.2 MBq, 0.2 mL, 200
verapamil) in the penile vein under isoflurane anesthesia. [18F]
1a was injected, and animals were transferred into the
microPET camera (microPET Focus 220, Siemens Medical
Solutions, Malvern, USA) causing a few minutes delay between
injection and start of a 60 min dynamic emission scan. All the
other radiopharmaceuticals were injected directly on the
camera following the start of a 30 min dynamic emission
scan. A transmission scan of 515 s with a 57Co point source was
performed for correction of attenuation and scatter by tissue
after an emission scan. Mice were terminated by cervical
dislocation. Several organs and tissues were excised and
weighed, and radioactivity was measured with a γ-counter.
Radioactivity accumulation in the organs was expressed as
standardized uptake value (SUV), using the following formula:
[tissue activity concentration (MBq/g)]/[injected dose
(MBq)/body weight (g)].
2.7. Metabolite Analysis. In vitro metabolism was
examined in human liver microsomes. A solution (4.0 mM,
14 μL) of 1a, 2, 3, or (R)-verapamil in DMSO was pipetted
into test tubes to a final concentration of 20.0 μM. Verapamil
was used as a reference compound to check if the microsomes
were working adequately because its metabolic pathways are
known.20 PBS (2.6 mL) and human liver microsomes (70.0 μL,
final protein concentration 0.5 mg/mL) were added, and tubes
were preincubated at 37 °C for 5 min. NADPH solution (20.0
mM, 140 μL) was freshly made in PBS and was added to tubes
in final concentration of 1 mM to start the reaction, as NADPH
is a cofactor for cytochrome P450 enzymes. The first sample
(400 μL) was taken immediately, and proteins were
precipitated by addition of acetonitrile + 0.1% formic acid
(800 μL). Other samples were taken at time points 15, 30, 45,
60, 90, and 120 min. Three control tubes were also incubated
for 120 min: one without microsomes, one without NADPH,
and one without test compound. In these tubes the volume of a
missing component was replaced by PBS. Sample tubes were
vortexed and centrifuged for 6 min at 12000 rpm. Supernatants
were analyzed by Xevo QTof UPLC/MS/MS system (Waters,
Milford, USA), mounted with ACQUITY UPLC BEH C18 1.7
μm column, H2O/CH3CN + 0.1% HCOOH as eluent at a flow
of 0.6 mL/min, ESI positive mode. Metabolynx XS 4.1 software
(Waters, Milford, USA) was used for the analysis of
Metabolites formed during the in vivo experiments were
analyzed from the plasma and brain tissue. Terminal arterial
whole blood samples (0.5 mL) were taken after the scans, and
they were centrifuged for 5 min at 6000 rpm to obtain plasma.
Plasma samples were precipitated by 2 volume addition of
acetonitrile and centrifuged, and 2.5−10 μL samples of
supernatant were applied on a thin layer chromatography
(TLC) plate (F-254 silica gel plates, Sigma-Aldrich, St. Louis,
USA). Metabolism in the brain was investigated by dissecting
half of the brain and homogenizing it with 0.9 mL of
acetonitrile. Brain homogenate was centrifuged, and super-
natant was applied on TLC. Plates were eluted with ethyl
acetate and methanol, ratio (v/v) 95:5 for [18F]1a and [18F]3
and ratio 99:1 for [18F]2 (Rf[18F]1a = 0.57, Rf[18F]2 = 0.69,
Rf[18F]3 = 0.41). Radioactivity of the eluted plates was analyzed
by phosphor storage imaging. Exposed screens were scanned
with a Cyclone phosphor storage system (PerkinElmer Life and
Analytical Science, Waltham, USA). Percentage of the intact
radiopharmaceutical and the formed metabolites were calcu-
lated by region of interest (ROI) analysis using OptiQuant
03.00 software (PerkinElmer, Waltham, USA).
2.8. PET Data Analysis. A 60 min scan, used for [18F]1a,
was separated into 21 time frames: 6 × 10, 4 × 30, 2 × 60, 1 ×
120, 1 × 180, 4 × 300, and 3 × 600 s. A 30 min scan used for
other radiopharmaceuticals consequently had the first 18 time
frames. Emission sinograms were normalized and corrected for
attenuation and radioactive decay. The sinograms were
iteratively reconstructed (two-dimensional ordered subsets
expectation maximization (OSEM2D) with Fourier rebinning,
4 iterations, and 16 subsets). The final data set consisted of 95
slices with a slice thickness of 0.8 mm and an in-plane image
matrix of 128 × 128 pixels. The voxel size was 0.9 mm × 0.9
mm × 0.8 mm. The Inveon Research Workplace software
version 4.0 (Siemens) was used for the data analysis. All the
frames were summed, and a PET image was coregistered with
an MRI template.21 Whole brain volume of interest (VOI)
based on the MRI template was generated. Radioactivity
concentration was converted to SUV values and plotted as a
time−activity curve (TAC).
2.9. Statistical Analysis. Differences between control
(FVB) and knockout animals were calculated for statistical
E
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
a
Scheme 2. Synthesis of 2
a
Reagents and conditions: (i) SOCl2 and Et3N; CH2Cl2 and 1.2% NaOH; 6,7-dimethoxytetrahydroisoquinoline; (ii) LiAlH4, THF; (iii) 2-
fluoroethyl tosylate, NaH, DMF.
a
Scheme 3. Synthesis of 3
a
Reagents and conditions: (i) CH3SO2Cl, Et3N, CH2Cl2, T = −10 °C; (ii) cyclopropylmagnesium bromide, 3 N HCl, THF; (iii) 6,7-dimethoxy-
1,2,3,4-tetrahydroisoquinoline, Na2CO3, DMF; (iv) H2, Pd/C (5%); (v) 2-fluoroethyl tosylate, NaH, DMF.
Table 1. Results of the Cell Experiments with Nonradioactive Compounds
compd
Papp BA/AB
P-gp EC50 (μM)
Bcrp EC50 (μM)
Mrp1 EC50 (μM)
ATPase activation
sigma-1 Ki (nM)
sigma-2 Ki (nM)
1a
2
9.5
7.9
6.1
2.1
15
>100
>100
>100
yes
yes
yes
>1000 (35%)
>1000 (28%)
>1000 (40%)
>1000 (25%)
>1000 (38%)
>1000 (44%)
0.54
0.35
>100
>100
3
significance using either one way analysis of variance (ANOVA)
with Bonferroni correction or two-sided unpaired Student’s t
test. A p-value of less than 0.05 was considered statistically
significant. IBM SPSS Statistics version 22 (Armonk, USA) was
used for the analysis.
compounds were transformed into mesylate or tosylate
derivatives. The leaving group was attempted to substitute
with fluoride by employing different reagents (NaF, tetrabuty-
lammonium fluoride), but unfortunately the reaction was
unsuccessful. Then, as a second approach, fluoroethyl tosylate
was prepared by reaction of fluoroethanol with toluensulphonyl
chloride. Phenol precursor was reacted with the fluoroethyl
tosylate in the presence of different bases (NaOH, KOH, NaH)
and solvents (toluene, DMF). Phenol precursors had low
reactivity and required the use of NaH and DMF for successful
synthesis.
3.2. Cell Experiments. Three different cell experiments
were performed with the nonradioactive compounds to
examine the substrate potential: apparent permeability experi-
ment in Caco-2 cells, Calcein-AM assay to determine selectively
the potency (EC50) toward each transporter, and ATP
depletion assay. The experimental setup is described in detail
elsewhere.18,24 A compound is defined as a substrate when it is
transported, and it activates ATPase. The ratio of drug
transport through Caco-2 monolayers in the basolateral-apical
3. RESULTS
3.1. Organic Synthesis. Compounds 1a, 2, and 3 were
synthesized as displayed in Schemes 1, 2, and 3, respectively.
Synthesis of compounds 1b, 4b−6b, and 7−14 has been
reported previously, and spectral data of these compounds were
identical to those previously described.17,22,23 Synthesis of
compound 1a was performed in the same conditions as the
nitro derivative 1b. The key intermediate, chloromethyloxazole
6a, was obtained by cyclization of fluorobenzamide 5a with
dichloroacetone. Compound 1a was obtained in a high yield
(74%) by condensation of 6a with 6,7-dimethoxytetrahydroi-
soquinoline. In the synthesis of fluoroethyl compounds 2 and 3,
different approaches were attempted. First, phenol precursors
were alkylated with bromoethanol. The resulting hydroxyethyl
F
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
Scheme 4. Radiosynthesis of [18F]1a
Scheme 5. Radiosynthesis of [18F]2 and [18F]3
Figure 1. Cerebral kinetics of the evaluated radiopharmaceuticals. Data are SUV-PET values for the entire brain, expressed as mean SEM (n = 3−
6). Time point 0 min is the start of the scan.
and apical-basolateral directions (Papp BA/AB) was higher than
the cutoff value of 2 for all the compounds (9.5, 7.9, and 6.1,
respectively, for 1a, 2, and 3) as presented in Table 1. EC50
values were determined in the Calcein-AM assay in Madin−
Darby Canine Kidney cells overexpressing selectively each
transporter. Potency was tested also against multidrug
resistance-associated protein 1 (Mrp1), which is another efflux
transporter. All compounds showed high affinity toward P-gp,
G
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
Figure 2. Biodistribution study of all the radiopharmaceuticals performed after the scans. Data are SUV expressed as mean SEM (n = 3−6).
Statistical differences (p < 0.05) are marked with a capped line.
most active being 3 (EC50 = 0.35 μM). Only 1a had some
affinity to Bcrp (15 μM), but none of the compounds were
found to have affinity to Mrp1. All the compounds activated
ATPase. Based on the results from all the experiments,
compounds were classified as P-gp substrates. Affinity of the
compounds to sigma-1 and sigma-2 receptors was also tested at
several concentrations in binding competition assays using
previously published methods, as our molecules and many of
the sigma ligands have a similar structure.25,26 The half maximal
inhibitory concentration IC50 could not be determined, and
consequently neither could inhibition constant Ki, because the
percentage of competition in all concentrations was less than
50%. Therefore, the percentage of competition at the highest
tested concentration (1 μM) is reported in parentheses for each
compound in Table 1. All compounds were found to be
inactive toward sigma-1 and sigma-2 receptors, the Ki values are
at least 100 times higher than the values for known sigma
ligands.27
[11C]Verapamil was produced in 39 min with a radiochemical
yield of 8.2 4.9%, specific activity of 34 16 GBq/μmol, and
over 99% radiochemical purity. Radiochemical yield is
calculated for the formulated product from the trapped activity
of [11C]CH4. Log D value as measured earlier was 2.7.28
3.4. PET Data and Biodistribution. In the [18F]1a scan
the peak of activity in the brain was missed, due to the delay
between radiopharmaceutical injection and start of the scan. On
average, the first 7.7 min were not recorded. However, it is clear
from the time−activity curves (Figure 1) that there is no
difference in the brain uptake between knockout and control
animals. In the biodistribution study (Figure 2), the uptake in
all the organs and tissues was also similar between the strains (p
> 0.05).
Scan time of other radiopharmaceuticals was reduced to 30
min, compared to 60 min for 1a, as it was found to be sufficient
to evaluate the cerebral uptake between the mice strains.
Injections were performed directly on the microPET camera to
avoid the loss of data. The cerebral kinetics of [18F]2 were not
significantly different between the strains (p = 0.092). In
addition, the shape of the time−activity curve was different than
expected. There is no initial peak of uptake, but instead the
curves are slowly rising until reaching a plateau after 10 min. In
the biodistribution data, no significant difference between the
strains was found. In the periphery, [18F]2 had only some
nonspecific uptake in all the collected organs and tissues.
Notable is the high uptake in plasma, which could refer to
binding in plasma proteins.
3.3. Radiolabeling. [18F]1a was synthesized via nucleo-
philic aromatic substitution reaction in 90 min with a total
radiochemical yield of 3.2
2.6% (Scheme 4). The
radiochemical purity was >96% and the specific activity 29
13 GBq/μmol. Radiochemical yield (decay corrected) is
calculated for the formulated product from the end of
bombardment (EOB) of [18F]fluoride. The measured log D
was 2.6. [18F]2 and [18F]3 were synthesized in a two-pot
method in 70−80 min (Scheme 5). They were produced with a
radiochemical yield of 11
4.7%. Specific activity for both
radiopharmaceuticals was >100 GBq/μmol. Radiochemical
purity was >98% for [18F]2 and >95% for [18F]3. Log D
value was measured as 2.9 for [18F]2 and 3.0 for [18F]3. (R)-
[18F]3 had almost the same maximum uptake in the Mdr1a/
b
(−/−)Bcrp1(−/−) mice as [11C]verapamil, but the excretion from
the brain was slower. Brain uptake of [18F]3 in the control mice
H
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
Figure 3. SUV-PET images of [18F]3 in control mouse (left), Mdr1a/b(−/−) mouse (middle), and Mdr1a/b(−/−)Bcrp1(−/−) mouse (right), sagittal
view. Vinci 4.24 software (Max Planck Institute, Cologne, Germany) was used to create the images.
was higher than for [11C]verapamil. The area under curve
(AUC) 0−30 min of [11C]verapamil in Mdr1a/b(−/−)Bcrp1(−/−)
mice was 3.7-fold higher than in the control mice. In
comparison with [18F]3, the difference was 2-fold. Maximum
knockout/control ratio with [18F]3 was reached in 15−20 min,
whereas for [11C]verapamil this was 5 min. There was no
statistically significant difference between the [18F]3 uptake in
Mdr1a/b(−/−)Bcrp1(−/−) and Mdr1a/b(−/−) knockout mice (p =
0.316), although the SUV values are higher for Mdr1a/
verapamil. Furthermore, baseline brain uptake of [18F]3 was
approximately 4-fold higher than for [11C]verapamil.
3.5. Metabolism. Phase I metabolism of the nonradioactive
compounds was investigated in the human liver microsomes, to
study the possible metabolites and metabolism rate. Control
tubes without microsomes and NADPH were included to see if
the compound metabolized/degraded without them and if the
microsomes were working adequately. Tubes without the test
compound were analyzed to detect impurities in the chemicals
or in the LC-MS system. No metabolites were observed in the
control tubes with any of the tested compounds.
In the experiment performed in human liver microsomes,
60% of the intact 1a was still present after 75 min.
Decomposition of the compound caused a formation of 6,7-
dimethoxy-1,2,3,4-tetrahydroisoquinoline, which was identified
by MS as a primary metabolite. In vivo plasma samples were
analyzed by radio-TLC. Majority of the metabolites were
hydrophilic (>85%, Rf = 0). After 75 min, 43 8.0% of the
parent radiopharmaceutical was still intact. Brain metabolism
data is not available.
b
(−/−)Bcrp1(−/−) mice. The area under the curve of Mdr1a/
b(−/−) mice constitutes about 86% of the AUC of Mdr1a/
(−/−)Bcrp1(−/−)mice. The examples of brain images with [18F]3
b
in each strain are displayed in Figure 3.
In the biodistribution study the uptake of [11C]verapamil was
significantly different between the strains in testes and brain at
time point 45 min after injection. Statistical differences between
control animals and Mdr1a/b(−/−)Bcrp1(−/−) knockouts in [18F]
3 biodistribution were found in the spleen, small intestine,
testes, and brain. Between Mdr1a/b(−/−) knockouts and
controls the statistical difference was found in whole blood,
plasma, liver, spleen, small intestine, and brain. In contrast to
the brain, the radioactivity uptake of [18F]3 in many P-gp
containing peripheral organs was actually higher in control
animals than in the knockouts.
Compound 2 was stable during the microsome study, and no
metabolites were found. In the light of in vivo data, this could be
due to the high protein binding. Only the supernatant was
analyzed, but the amount of the compound in the precipitate is
In addition to TACs, we calculated brain-to-plasma radio-
activity ratios from the biodistribution data. Therefore, the
values represent only one time point after the scan (Table 2).
Brain-to-plasma radioactivity ratios of [18F]3 suggest that
radioligand is P-gp selective and is not transported by Bcrp.
The difference in brain uptake between control and P-gp
knockout was 4.1-fold for [18F]3 and 7.1-fold for [11C]-
not known. At 45 min, 78
10% of the parent [18F]2 was
intact in the plasma and 69 10% in the brain, based on the
radio-TLC. All the observed metabolites were hydrophilic (Rf =
0).
Demethylation and defluoroethylation were observed for 3 in
microsome incubations. After 45 min, 80% of the parent was
still intact. In vivo at the same time point the number was 69
11% in the plasma and 96 0.8% in the brain. All the found
metabolites were hydrophilic in the nature (Rf = 0−0.3).
In vivo [11C]verapamil metabolite analysis was not possible
due to the low activity in the plasma and brain. However,
metabolism of (R)-[11C]verapamil is known in rats.20
According to that study, at 30 min postinjection, 47% of the
parent radiopharmaceutical was intact in plasma, and at 60 min,
28%. In the brain at 30 min, 68% of the intact radio-
pharmaceutical was found, and at 60 min, 47%. Typical
metabolites for verapamil are O- and N-demethylated
compounds, as well as an N-dealkylated one.20 These
metabolites were found also in our microsome experiment
where about 60% of the parent compound was intact at 45 min.
Table 2. Brain-to-Plasma Ratios of All Groups; Data Are
Presented as Mean SD
groups
brain-to-plasma
[11C]verapamil control
[11C]verapamil Mdr1a/b(−/−)Bcrp1(−/−)
[18F]1a control
[18F]1a Mdr1a/b(−/−)Bcrp1(−/−)
[18F]2 control
[18F]2 Mdr1a/b(−/−)Bcrp1(−/−)
[18F]3 control
0.32 0.1
2.3 0.2
0.45 0.3
0.38 0.1
0.37 0.04
0.54 0.1
1.3 0.2
5.6 0.8
4.2 1.1
[18F]3 Mdr1a/b(−/−)
[18F]3 Mdr1a/b(−/−)Bcrp1(−/−)
I
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
4. DISCUSSION
Article
Both precursors of 2 and 3 (9 and 14) have been previously
labeled with carbon-11.23,30 The carbon-11 analogue of 2,
[11C]MC113, was characterized in vitro as an inhibitor but in
vivo was weakly transported by P-gp. Uptake in the Mdr1a/
b(−/−) mice was higher than in the control mice. In high and
low P-gp expressing tumors [11C]MC113 also failed to show
the difference. Change of the methyl group into a fluoroethyl in
2 affected the affinity negatively in vivo based on the results
presented in this Article, although in vitro they were almost the
same (MC113, 0.6 μM; 2, 0.54 μM). Instead of affinity, a
reason for this might be the better metabolic stability of
[11C]MC113.
Three novel PET radiopharmaceuticals designed for imaging P-
glycoprotein function were synthesized and evaluated in control
and knockout mice. All of the compounds were labeled with
fluorine-18 to increase their applicability. Compounds were
characterized in vitro as substrates and were compared in vivo to
the best known P-gp substrate [11C]verapamil.
Low radiochemical yield of [18F]1a (3.2%) is limiting its use.
In order to improve the radiochemical yield, several different
reaction temperatures and times were investigated. Microwave
and microfluidic synthesis modules were also attempted to
conduct the labeling, but the radiochemical yield remained low.
The oxazole ring in the para-position is not a very strong
activator for the nitro-group of the precursor. However, to
conduct the current studies, the radiochemical yield was
sufficient. Other radiopharmaceuticals were synthesized with a
decent radiochemical yield of about 10%.
The carbon-11 analogue of [18F]3, [11C]MC266, was
evaluated in normal rats and in rats pretreated with
cyclosporine A (CsA).23 Results were comparable to [18F]3
studies in mice, as described in this Article. Cyclosporine A
treatment increased the brain uptake of [11C]MC266 2.2-fold
(AUC0−60 min). CsA also increased the cerebral distribution
volume 3-fold. Except brain, CsA treatment did not affect the
radioactivity concentration in any peripheral organ.
As expected for a substrate, [18F]3 had higher uptake in the
transporter knockout mice compared to the control mice.
Uptake in the Mdr1a/b(−/−) knockout was a little bit lower than
in the Mdr1a/b(−/−)Bcrp1(−/−) knockout, indicating that Bcrp is
involved in the transport. However, the difference in the time−
activity curves was not statistically significant. In vitro the Bcrp
affinity was not observed. The higher uptake in Mdr1a/
5. CONCLUSION
In search for a new 18F-substrate radiopharmaceutical for
imaging functional changes of P-gp at the BBB, we evaluated
three novel compounds: [18F]1a, [18F]2, and [18F]3. Due to the
good metabolic stability and brain uptake, [18F]3 is the most
promising radiopharmaceutical for P-gp PET imaging. Further
quantitative in vivo studies in rats are needed to assess the
cerebral kinetics of the compound.
b
(−/−)Bcrp1(−/−) mice may be due to the cooperative lack of
both P-gp and Bcrp, and the pure Bcrp effect would be better
investigated with Bcrp1(−/−) knockout. This is in contrast to
(R)-[11C]verapamil, which has been found to be transported
only by P-gp and not by Bcrp or Mrp1 at the BBB.29
Uptake in control mice was higher for [18F]3 compared to
(R)-[11C]verapamil, which could be advantageous in the cases
where P-gp is overexpressed at the BBB. Upon up-regulation of
P-gp, a very low PET signal can be expected, and since [18F]3
shows higher brain uptake under normal physiological
condition, its PET signal might be significantly larger than
that of (R)-[11C]verapamil, assuming that the radioactivity
uptake in controls in P-gp mediated.
AUTHOR INFORMATION
Corresponding Author
*E-mail: g.luurtsema@umcg.nl. Tel: +31503612686.
■
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
[18F]1a and [18F]2 failed to show any substrate activity for P-
gp in vivo, although these compounds were selected from in
vitro assays as potential candidates. Neither of them showed any
difference in the brain uptake between Mdr1a/
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by Dutch Technology Foundation
STW (project number 11741) and ZonMW (project number
91111007, UPLC/TOF/MS). The authors would like to
acknowledge Chantal Kwizera for the help in development of
the labeling for [18F]1a and Marcel Benadiba for the animal
work with this radiopharmaceutical. We would also like to
thank Marianne Schepers for the UPLC/MS/MS analysis and
b
(−/−)Bcrp1(−/−)and control mice. In addition, [18F]2 had high
retention in plasma, which could indicate strong plasma protein
binding. Compound 1a has a 6-fold higher EC50 value than 3,
which could explain the poor in vivo results. Because 2 has an
EC50 value in the same range as 3, its substrate activity cannot
explain this result. However, it should be noted that metabolism
might play an important role in vivo.
Jurgen Sijbesma for the help in animal experiments.
̈
Compound 3 was shown to be metabolically more stable
than verapamil. In vivo metabolites were analyzed in plasma and
brain, and in vitro metabolism was investigated in liver
microsomes. Although liver microsomes were from a human
source and in vivo experiments were performed in mice,
metabolism rate in microsomes was quite predictable for the in
vivo situation. More than 96% of the parent radiopharmaceut-
ical [18F]3 was still intact in the brain after 45 min, whereas
(R)-[11C]verapamil produces much more brain entering
metabolites, as was found previously in rats.20 The in vivo
stability of a radiopharmaceutical is of importance because PET
measures only total radioactivity and cannot distinguish
radioactive metabolites from the parent compound.
REFERENCES
■
(1) Giacomini, K. M.; Huang, S. Transporters in Drug Development
and Clinical Pharmacology. Clin. Pharmacol. Ther. 2013, 94, 3−9.
(2) Chu, X.; Bleasby, K.; Evers, R. Species Differences in Drug
Transporters and Implications for Translating Preclinical Findings to
Humans. Expert Opin. Drug Metab. Toxicol. 2013, 9, 237−252.
(3) Higgins, C. F.; Linton, K. J. The ATP Switch Model for ABC
Transporters. Nat. Struct. Mol. Biol. 2004, 11, 918−927.
(4) Graff, C. L.; Pollack, G. M. Drug Transport at the Blood-brain
Barrier and the Choroid Plexuse. Curr. Drug Metab. 2004, 5, 95−108.
(5) Van Assema, D. M. E.; Lubberink, M.; Bauer, M.; van der Flier,
W. M.; Schuit, R. C.; Windhorst, A. D.; Comans, E. F. I.; Hoetjes, N.
J.; Tolboom, N.; Langer, O.; Muller, M.; Scheltens, P.; Lammertsma,
̈
J
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics
Article
A. a; van Berckel, B. N. M. Blood-Brain Barrier P-Glycoprotein
Function in Alzheimer’s Disease. Brain 2012, 135, 181−189.
(6) Deo, A. K.; Borson, S.; Link, J. M.; Domino, K.; Eary, J. F.; Ke, B.;
Richards, T. L.; Mankoff, D. a; Minoshima, S.; O’Sullivan, F.; Eyal, S.;
Hsiao, P.; Maravilla, K.; Unadkat, J. D. Activity of P-Glycoprotein, a B-
Amyloid Transporter at the Blood-Brain Barrier, Is Compromised in
Patients with Mild Alzheimer Disease. J. Nucl. Med. 2014, 55, 1106−
1111.
(7) Bartels, A.; Willemsen, A. T. M.; Kortekaas, R.; de Jong, B. M.; de
Vries, R.; de Klerk, O.; van Oostrom, J. C. H.; Portman, A.; Leenders,
K. L. Decreased Blood-Brain Barrier P-Glycoprotein Function in the
Progression of Parkinson ’s Disease, PSP and MSA. J. Neural Transm.
2008, 115, 1001−1009.
(8) Langer, O.; Bauer, M.; Hammers, A.; Karch, R.; Pataraia, E.;
Koepp, M. J.; Abrahim, A.; Luurtsema, G.; Brunner, M.; Sunder-
plassmann, R.; Zimprich, F.; Joukhadar, C.; Gentzsch, S.; Dudczak, R.;
Kletter, K.; Markus, M.; Baumgartner, C. Pharmacoresistance in
Epilepsy: A Pilot PET Study with the P-Glycoprotein Substrate R-
[11C]verapamil. Epilepsia 2007, 48, 1774−1784.
SAR Studies on Tetrahydroisoquinoline Derivatives. Bioorg. Med.
Chem. 2008, 16, 362−373.
(19) Wegman, T. D.; Maas, B.; Elsinga, P. H.; Vaalburg, W. An
Improved Method for the Preparation of [11C]verapamil. Appl. Radiat.
Isot. 2002, 57, 505−507.
(20) Luurtsema, G.; Molthoff, C. F. M.; Schuit, R. C.; Windhorst, A.
D.; Lammertsma, A. A.; Franssen, E. J. F. Evaluation of (R)-
[11C]verapamil as PET Tracer of P-Glycoprotein Function in the
Blood-Brain Barrier: Kinetics and Metabolism in the Rat. Nucl. Med.
Biol. 2005, 32, 87−93.
(21) Ma, Y.; Hof, P. R.; Grant, S. C.; Blackband, S. J.; Bennett, R.;
Slatest, L.; McGuigan, M. D.; Benveniste, H. A Three-Dimensional
Digital Atlas Database of the Adult C57BL/6J Mouse Brain by
Magnetic Resonance Microscopy. Neuroscience 2005, 135, 1203−1215.
(22) Colabufo, N. A.; Berardi, F.; Cantore, M.; Perrone, M. G.;
Contino, M.; Inglese, C.; Niso, M.; Perrone, R.; Azzariti, A.; Simone,
G. M.; Paradiso, A. 4-Biphenyl and 2-Naphthyl Substituted 6,7-
Dimethoxytetrahydroisoquinoline Derivatives as Potent P-Gp Modu-
lators. Bioorg. Med. Chem. 2008, 16, 3732−3743.
(23) Van Waarde, A.; Ramakrishnan, N. K.; Rybczynska, A. A.;
Elsinga, P. H.; Berardi, F.; de Jong, J. R.; Kwizera, C.; Perrone, R.;
Cantore, M.; Sijbesma, J. W. A.; Dierckx, R. A.; Colabufo, N. A.
Synthesis and Preclinical Evaluation of Novel PET Probes for P-
Glycoprotein Function and Expression. J. Med. Chem. 2009, 52, 4524−
4532.
(24) Colabufo, N. A.; Berardi, F.; Perrone, R.; Rapposelli, S.;
Digiacomo, M.; Vanni, M. 2-[(3-Methoxyphenylethyl) Phenoxy]-
Based ABCB1 Inhibitors: Effect of Different Basic Side-Chains on
Their Biological Properties. J. Med. Chem. 2008, 51, 7602−7613.
(25) Abate, C.; Pati, M. L.; Contino, M.; Colabufo, N. A.; Perrone,
R.; Niso, M.; Berardi, F. From Mixed Sigma-2 receptor/P-
Glycoprotein Targeting Agents to Selective P-Glycoprotein Modu-
lators: Small Structural Changes Address the Mechanism of
Interaction at the Efflux Pump. Eur. J. Med. Chem. 2015, 89, 606−615.
(26) Matsumoto, R. R.; Bowen, W. D.; Tom, M. A.; Vo, V. N.;
Truong, D. D.; Costa, B. R. Characterization of Two Novel σ Receptor
Ligands: Antidystonic Effects in Rats Suggest σ Receptor Antagonism.
Eur. J. Pharmacol. 1995, 280, 301−310.
(9) Colabufo, N. A.; Berardi, F.; Perrone, M. G.; Capparelli, E.;
Cantore, M.; Inglese, C.; Perrone, R. Substrates, Inhibitors and
Activators of P-Glycoprotein: Candidates for Radiolabeling and
Imaging Perspectives. Curr. Top. Med. Chem. 2010, 10, 1703−1714.
(10) Luurtsema, G.; Molthoff, C. F. M.; Windhorst, A. D.; Smit, J.
W.; Keizer, H.; Boellaard, R.; Lammertsma, A. A.; Franssen, E. J. F.
(R)- and (S)-[11C]verapamil as PET-Tracers for Measuring P-
Glycoprotein Function: In Vitro and in Vivo Evaluation. Nucl. Med.
Biol. 2003, 30, 747−751.
(11) Bauer, M.; Zeitlinger, M.; Karch, R.; Matzneller, P.; Stanek, J.;
Jager, W.; Bohmdorfer, M.; Wadsak, W.; Mitterhauser, M.; Bankstahl,
̈
̈
J. P.; Loscher, W.; Koepp, M.; Kuntner, C.; Muller, M.; Langer, O.
̈
̈
Pgp-Mediated Interaction between (R)-[11C]verapamil and Tariquidar
at the Human Blood-Brain Barrier: A Comparison with Rat Data. Clin.
Pharmacol. Ther. 2012, 91, 227−233.
(12) Bauer, M.; Karch, R.; Neumann, F.; Abrahim, A.; Wagner, C. C.;
Kletter, K.; Muller, M.; Zeitlinger, M.; Langer, O. Age Dependency of
̈
Cerebral P-Gp Function Measured with (R)-[11C]verapamil and PET.
Eur. J. Clin. Pharmacol. 2009, 65, 941−946.
(27) Van Waarde, A.; Rybczynska, A. A.; Ramakrishnan, N. K.;
Ishiwata, K.; Elsinga, P. H.; Dierckx, R. A. J. O. Potential Applications
for Sigma Receptor Ligands in Cancer Diagnosis and Therapy.
Biochim. Biophys. Acta 2014, DOI: 10.1016/j.bbamem.2014.08.022.
(28) Knutson, T. W.; Knutson, L.; Jansson, B.; Lennerna, H. The
Effect of Ketoconazole on the Jejunal Permeability and CYP3A
Metabolism of (R/S)-Verapamil in Humans. J. Clin. Pharmacol. 1999,
48, 180−189.
(13) Bankstahl, J. P.; Bankstahl, M.; Romermann, K.; Wanek, T.;
̈
Stanek, J.; Windhorst, A. D.; Fedrowitz, M.; Erker, T.; Muller, M.;
̈
̈
Loscher, W.; Langer, O.; Kuntner, C. Tariquidar and Elacridar Are
Dose-Dependently Transported by P-Glycoprotein and Bcrp at the
Blood-Brain Barrier: A Small-Animal Positron Emission Tomography
and in Vitro Study. Drug Metab. Dispos. 2013, 41, 754−762.
(14) Kannan, P.; Telu, S.; Shukla, S.; Ambudkar, S. V.; Pike, V. W.;
Halldin, C.; Gottesman, M. M.; Innis, R. B.; Hall, M. D. The “Specific”
P-Glycoprotein Inhibitor Tariquidar Is Also a Substrate and an
Inhibitor for Breast Cancer Resistance Protein (BCRP/ABCG2). ACS
Chem. Neurosci. 2011, 2, 82−89.
(29) Romermann, K.; Wanek, T.; Bankstahl, M.; Bankstahl, J. P.;
̈
Fedrowitz, M.; Muller, M.; Loscher, W.; Kuntner, C.; Langer, O. (R)-
̈
̈
[11C]verapamil Is Selectively Transported by Murine and Human P-
Glycoprotein at the Blood-Brain Barrier, and Not by MRP1 and
BCRP. Nucl. Med. Biol. 2013, 40, 873−878.
(15) Kawamura, K.; Yamasaki, T.; Konno, F.; Yui, J.; Hatori, A.;
Yanamoto, K.; Wakizaka, H.; Ogawa, M.; Yoshida, Y.; Nengaki, N.;
Fukumura, T.; Zhang, M.-R. Synthesis and in Vivo Evaluation of 18F-
Fluoroethyl GF120918 and XR9576 as Positron Emission Tomog-
raphy Probes for Assessing the Function of Drug Efflux Transporters.
Bioorg. Med. Chem. 2011, 19, 861−870.
(30) Mairinger, S.; Wanek, T.; Kuntner, C.; Doenmez, Y.; Strommer,
S.; Stanek, J.; Capparelli, E.; Chiba, P.; Muller, M.; Colabufo, N. A.;
̈
Langer, O. Synthesis and Preclinical Evaluation of the Radiolabeled P-
Glycoprotein Inhibitor [11C]MC113. Nucl. Med. Biol. 2012, 39, 1219−
1225.
(16) Dorner, B.; Kuntner, C.; Bankstahl, J. P.; Wanek, T.; Bankstahl,
̈
M.; Stanek, J.; Mullauer, J.; Bauer, F.; Mairinger, S.; Loscher, W.;
̈
̈
Miller, D. W.; Chiba, P.; Muller, M.; Erker, T.; Langer, O.
̈
Radiosynthesis and in Vivo Evaluation of 1-[18F]fluoroelacridar as a
Positron Emission Tomography Tracer for P-Glycoprotein and Breast
Cancer Resistance Protein. Bioorg. Med. Chem. 2011, 19, 2190−2198.
(17) Colabufo, N. A.; Berardi, F.; Perrone, M. G.; Cantore, M.;
Contino, M.; Inglese, C.; Niso, M.; Perrone, R. Multi-Drug-Resistance-
Reverting Agents: 2-Aryloxazole and 2-Arylthiazole Derivatives as
Potent BCRP or MRP1 Inhibitors. ChemMedChem 2009, 4, 188−195.
(18) Colabufo, N. A.; Berardi, F.; Cantore, M.; Perrone, M. G.;
Contino, M.; Inglese, C.; Niso, M.; Perrone, R.; Azzariti, A.; Simone,
G. M.; Porcelli, L.; Paradiso, A. Small P-Gp Modulating Molecules:
K
DOI: 10.1021/mp5008103
Mol. Pharmaceutics XXXX, XXX, XXX−XXX