Synthesis and in vitro evaluation of 18F-β-carboline alkaloids as PET ligands

Article abstract of DOI:10.1002/jlcr.1519

A one-step ¹⁸F-labelling strategy was used to prepare four ¹⁸F-labelled analogues of 7-methoxy-1-methyl-9H-β-carboline (harmine): 7-(2-[¹⁸F]fluoroethoxy)-1-methyl-9H-β-carboline (5), 7-(3-[¹⁸F]fluoro-propoxy)-1-methyl-9H-β-carboline (6), 7-[2-(2-[¹⁸F]fluoroethoxy)ethoxy]-1-methyl-9H-β-carboline (7), and 7-{2-[2-(2-[¹⁸F]fluoroethoxy)ethoxy]-ethoxy}-1-methyl-9H-β- carboline (8). These were synthesized as potential positron emission tomography ligands for monoamine oxidase A (MAO-A). A solution of pure labelled compound in buffer was obtained in < 70 min from end of radionuclide production, with a decay-corrected yield of up to 23%. The average specific binding to MAO-A in rat brain, determined by autoradiography experiments, was highest for compounds 7 and 8 (89±2 and 96±1%, respectively), which was obtained at <1 nM radioligand concentration. Copyright

Full text of DOI:10.1002/jlcr.1519

Research Article
Received 18 December 2007,
Revised 14 March 2008,
Accepted 25 March 2008
Published online in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1519
18
Synthesis and in vitro evaluation of F-b-
carboline alkaloids as PET ligands
˚
a
a
b
c
Elisabeth Blom, Farhad Karimi, Olof Eriksson, Hakan Hall, and Bengt
a,c,d
Ã
˚
Langstrom
¨
A one-step 18F-labelling strategy was used to prepare four 18F-labelled analogues of 7-methoxy-1-methyl-9H-b-carboline
(harmine): 7-(2-[18F]fluoroethoxy)-1-methyl-9H-b-carboline (5), 7-(3-[18F]fluoro-propoxy)-1-methyl-9H-b-carboline (6), 7-[2-
(2-[18F]fluoroethoxy)ethoxy]-1-methyl-9H-b-carboline (7), and 7-{2-[2-(2-[18F]fluoroethoxy)ethoxy]-ethoxy}-1-methyl-9H-b-
carboline (8). These were synthesized as potential positron emission tomography ligands for monoamine oxidase A (MAO-
A). A solution of pure labelled compound in buffer was obtained in o70 min from end of radionuclide production, with a
decay-corrected yield of up to 23%. The average specific binding to MAO-A in rat brain, determined by autoradiography
experiments, was highest for compounds 7 and 8 (8972 and 9671%, respectively), which was obtained at o1 nM
radioligand concentration.
Keywords: harmine analogues; nucleophilic 18F-fluorination; one-step labelling; b-carboline alkaloids; MAO-A
[11C]methylprop-2-yn-1-amine ([11C]CLG),7
and
N-[3-(2,4-
Introduction
dichlorophenoxy)-2-[18F]fluoropropyl]-N-methylprop-2-yn-1-
amine ([18F]CLG)15 have previously been used and explored as
PET tracers for characterization of MAO-A enzyme. Harmine was
chosen as the basis of the 18F-labelled analogues because
attaching a side chain to its phenol group does not significantly
alter its chemical properties, especially the pKa (Table 2).
The syntheses of 7-(2-[18F]fluoroethoxy)-1-methyl-9H-b-carbo-
line (5), 7-(3-[18F]fluoro-propoxy)-1-methyl-9H-b-carboline (6),
Two b-carbolines, harmine and harmaline, which are based on an
indole nucleus fused with a pyridine ring, can be isolated from
Penganum harmala.1 Harmine derivatives have various biological
activities, such as interaction with benzodiazepine receptors,2
inhibition of cyclin-dependent kinase,3 and inhibition of mono-
amine oxidase (MAO),4
a membrane-bound mitochondrial
enzyme5 with two subtypes, MAO-A and -B6. [11C]Harmine has
been used and validated in vivo in monkey brain to bind to MAO-
A.7 This tracer was also used in humans in a competition study
with other MAO-A active drugs,8 and its binding to MAO-A has
been quantified in the human brain.9 [11C]Harmine has as well
been used in a dual tracer positron emission tomography (PET)
study together with [11C]raclopride in pig10 and its biodistribution
and radiation dosimetry evaluated in baboons.11 Recently we
observed that the high expression of MAO-A could be exploited
for characterization of neuroendocrine gastroenteropancreatic
tumours by using [11C]harmine as a tracer.12 Thus, four 18F-
labelled analogues of harmine (Scheme 1) have been synthesized
to explore their potential as selective MAO-A tracers using PET,
taking advantage of the longer half-life of 18F (110min) compared
with 11C (20.3min). The 18F-label was placed in the side chain, to
investigate the biological properties conferred by changes in its
chemical properties (i.e. by modifying lipophilicity using a
polyethylene glycol (PEG) strategy13). The binding properties of
the four 18F-labelled analogues were evaluated using autoradio-
7-[2-(2-[18F]fluoroethoxy)ethoxy]-1-methyl-9H-b-carboline
(7),
and 7-{2-[2-(2-[18F]fluoroethoxy)ethoxy]ethoxy}-1-methyl-9H-b-
carboline (8) as the target compounds from the corresponding
2-[(1-methyl-9H-b-carbolin-7-yl)oxy]ethyl 4-methylbenzenesulfo-
nate (1), 3-[(1-methyl-9H-b-carbolin-7-yl)oxy]propyl 4-methyl-
benzenesulfonate (2), 7-[2-(2-bromoethoxy)ethoxy]-1-methyl-
9H-b-carboline (3), and 7-{2-[2-(2-chloroethoxy)ethoxy]ethoxy}-
1-methyl-9H-b-carboline (4) are outlined in Scheme 1.
The labelled compounds 5–8 were synthesized from the
precursors 1–4 using no-carrier-added [18F]fluoride in a one-step
nucleophilic substitution (Scheme 1). Heating the reaction
mixture containing a precursor and [K/K2.2.2.]118Fꢀ in 400 mL
N,N-dimethylformamide (DMF) at 1501C for 15 min gave the
aDepartment of Biochemistry and Organic Chemistry, Uppsala University, Box
576, Husargatan 3, BMC, S-751 23 Uppsala, Sweden
bDepartment of Radiology, Oncology and Clinical Immunology, Division of
graphy on sections from the rat brain, an organ that is known to Radiology, Rudbeck Laboratory, Uppsala University, S-751 85 Uppsala, Sweden
have high density of MAO-A.7
cUppsala Applied Science Lab, GEHC, Box 967, S-751 09 Uppsala, Sweden
dUppsala Imanet, GEHC, Box 967, S-751 09 Uppsala, Sweden
Results and discussion
˚
*Correspondence to: Bengt Langstro¨m, Uppsala Imanet, GEHC, Box 967, S-751
09 Uppsala, Sweden.
E-mail: Bengt.Langstrom@biorg.uu.se
A number of tracers such as 7-[11C]methoxy-1-methyl-9H-b-
carboline ([11C]HAR),7,14 N-[3-(2,4-dichlorophenoxy)propyl]-N-
J. Label Compd. Radiopharm 2008, 51 277–282
Copyright r 2008 John Wiley & Sons, Ltd.

E. Blom et al.
18F-
N
N
O
O
R1
N
H
R2
N
H
1: R1 = (CH2)2OTs
2: R1 = (CH2)3OTs
5: R2 = (CH2)218F
6: R2 = (CH2)318F
7: R2 = (CH2)2O(CH2)218F
3: R1 = (CH2)2O(CH2)2Br
4: R1 = (CH2)2O(CH2)2O(CH2)2Cl
8: R2 = (CH2)2O(CH2)2O(CH2)218F
Scheme 1. 18F-Labelling reaction to yield compounds 5–8.
Table 1. Radiochemical yields and specific activity in
labelling reactions
coherence nuclear magnetic resonance (1H–13C HMBC NMR)
experiments to confirm that the precursors were O-alkylated and
not N-alkylated. It should be pointed out that 1–4 are not stable
when stored at -201C for more than approximately 1 month.
Compounds 3 and 4 are stable for 6 months when stored at -701C,
but 1 and 2 are not. The order of magnitude of the logP values
calculated using ACD-labs17 for compounds 5–8 were in agree-
ment with the retention times of the respective compounds using
a standardized analytical HPLC program (Table 2).
Several approaches for the preparation of 1, the precursor to
the ethyl analogue 5, were explored. Three different bases,
pyridine, potassium tert-butoxide, and caesium carbonate, were
used. Caesium carbonate was successful in creating the
phenoxide, and was therefore used in all precursor syntheses.
In vitro autoradiography experiments were carried out to
investigate the binding properties of labelled compounds 5–8
to MAO-A in rat brain (Figure 1). All compounds showed some
degree of specific binding to the cerebral cortex and striatum;
both areas have high densities of the target enzyme MAO-A. The
uptake was similar in the different brain regions, with slightly
higher binding observed in the anterior cingulate cortex and
dorsal striatum. The average percentage of specific binding was
highest for compounds 7 and 8, as illustrated in Table 3. For
comparison, [11C]HAR has specific binding of 78–86%.14
The 18F label was placed in different side chains in order to
obtain a set of compounds with a range of lipophilicities. PEG
segments of different lengths were used in compounds 7 and 8
to give different lipophilicities13 relative to the ethyl and propyl
analogues 5 and 6. The lower lipophilicity of 7 and 8 (Table 2)
might explain their higher specific binding.
Compound
RCY (%)a
Spec. act. (GBq/mmol)b
5
6
7
8
2373 (n = 4)
1072 (n = 2)
1273 (n = 5)
1474 (n = 4)
6057110 (n = 4)
744730 (n = 2)
5087160 (n = 5)
4407100 (n = 2)
aIsolated decay-corrected radiochemical yield, calculated
from the amount of radioactivity at the start of synthesis
and radioactivity of LC purified product, n = number of
experiments.
bRatio of radioactivity to amount of substance at end of
synthesis, n = number of experiments.
corresponding labelled product. The crude product was purified
by semi-preparative high-performance liquid chromatography
(HPLC). The decay-corrected radiochemical yield was in the
range 10–23% (Table 1). Compound 5 was prepared with the
highest radiochemical yield. The unlabelled reference
substances to compounds 5–8 were synthesized by reacting
the corresponding precursor with tetrabutylammonium fluoride
in tetrahydrofuran (THF) at 901C. These were used for
identification in the analysis of the 18F-labelled compounds in
all the liquid chromatographic (LC) runs. The identities of
compounds 5, 7, and 8 were confirmed by liquid chromato-
graphy mass spectrometry (LC-MS) analysis. In
a typical
experiment (compound 7) starting with 7.60 GBq of [18F]fluor-
ide, 0.93 GBq of purified product was obtained within 70 min
from end of radionuclide production. Specific activity was
determined for compounds 5–8 as illustrated in Table 1. The
radiochemical purity exceeded 98% in all labelling reactions. In
an attempt to increase the radiochemical yield tert-butanol was
added to the labelling reaction,16 but in this case it actually
decreased the yield.
Compounds 7 and 8 were chosen for metabolite analysis
because of their high specific binding and the relative stability of
the corresponding precursors. The fraction of unmetabolized
tracer in rat blood plasma was 26 and 52% after 5 min, and 4 and
12% after 30 min for compounds 7 and 8, respectively. The use
of a pegylated side chain in which the 18F label was placed
might also decrease the metabolism, but this has not been
investigated in this study.
The precursors 1–4 were synthesized from 1-methyl-9H-b-
carbolin-7-ol (harmol) and either the corresponding alkyl di(p-
toluenesulfonate), 1-bromo-2-(2-bromoethoxy)ethane, or 1-chloro-
2-[2-(2-chloroethoxy)ethoxy]ethane (Scheme 2). These were com-
bined with caesium carbonate in dry DMF at 01C or, in some cases,
at -781C. After the reagents were added, the temperature was
slowly raised to room temperature. Attempts to perform the
reactions at higher temperature resulted in very poor yields, or No-carrier-added aqueous [18F]fluoride was produced from
even in exclusive formation of side products. The structures of 1–4 water 95% enriched in 18O (Rotem Industries Ltd., Israel or
were elucidated using 1H–13C heteronuclear multiple bond Taiyo Nippon Sanso Corporation) by the nuclear reaction
Experimental
Radiosyntheses
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 277–282

E. Blom et al.
Table 2. Retention times on analytical HPLC and calculated log P and pKa values for compounds 5–8
c
pKa
Compound
log Pa
Retention timeb (min)
Indole
Pyridine
5
6
7
8
3.4071.08
3.6971.07
3.0071.11
2.6471.14
7.6
8.1
7.3
7.2
15.6170.40
15.5270.40
15.5770.40
15.5770.40
8.1270.40
8.0470.40
8.0870.40
8.0970.40
aCalculated using ACD-labs.17
bRetention time on analytical HPLC column.
cApproximated apparent pKa values of indole and pyridine nitrogen, calculated using ACD-labs.17
Cs2CO3
N
N
HO
O
R1
DMF
N
N
H
H
1: R1 = (CH2)2OTs
2: R1 = (CH2)3OTs
OTs
OTs
TsO
TsO
O
O
3: R1 = (CH2)2O(CH2)2Br
Br
Br
Cl
4: R1 = (CH2)2O(CH2)2O(CH2)2Cl
O
Cl
Scheme 2. Synthesis of precursors 1–4.
Table 3. Percent specific binding of the four harmine
analogues determined from in vitro autoradiography
Specific
binding
(%)
Ligand
concentration
(nM)
Compound
na
5
6
7
8
4876
4878
8972
9671
0.21270.064
0.18870.081
0.16170.059
0.17070.063
12
7
10
10
an = number of experiments.
transferred from the cyclotron target by HPLC pump and
trapped on a QMA filter (ABX, advanced biochemical com-
pounds, pre-conditioned Sep-PAKs, light QMA cartridge with
CO23ꢀ as counter ions, Radeberg). The QMA filter was purged
with helium for 3 min and thereafter the [18F]fluoride was
released with a 2 mL solution of 96:4 (by volume, total volume
12 mL) acetonitrile–water mixture containing 55.9 mg of Kryp-
tofix 2.2.2 (K2.2.2) and 12.7 mg K2CO3. The eluted 18F/Kryptofix/
K2CO3 solution was dried under N2 (g) at 1101C and then with
2 ꢁ 1 mL dry acetonitrile. Synthia,18 an automated synthesis
system, was used for the handling of the reagents. LC analysis
was performed with a VWR Hitachi Pump L-2130 and a VWR
Hitachi UV Detector L-2400 UV detector in series with a b1-flow
detector. A Discoverys C18, Supelco, 25 cm ꢁ 4.6 mm, 5 mm
HPLC column was used. The following mobile phases were used:
25 mM KH2PO4 in water (A) and acetonitrile/water 50:7 (B).
Program: 10–100% (B) over 5 min, then 100% for 10 min, flow
rate 1.5 mL/min. For semi-preparative LC an ACE 5C18HL,
Scantec Lab, 250 mm ꢁ 10 mm, 5 mm column was used at a
flow rate of 5 mL/min. Acetonitrile (25%) and 0.05 M ammonium
formate pH 3.5 (75%) were used as eluent. The identities of the
Figure 1. Colour-coded images of total and non-specific binding of the four
harmine analogues 5–8.
18O(p,n)18F using a Scanditronix MC-17 cyclotron at Uppsala
Imanet. The produced solution of [18F]fluoride in water was
J. Label Compd. Radiopharm 2008, 51 277–282
Copyright r 2008 John Wiley & Sons, Ltd.
www.jlcr.org

E. Blom et al.
1
18F-labelled compounds were confirmed by co-injection of methanol 10:1) to yield 1 as a yellow oil (0.246 g, 8%). H NMR
isotopically unmodified reference substances in all the LC runs. (400 MHz, CD3OD, 251C) d = 8.08 (d, 3JH,H = 5.5 Hz, 1H), 7.90 (dd,
3
LC-MS analyses of 18F-labelled compounds were performed 5JH,H = 0.5, JH,H = 8.7 Hz, 1H), 7.76 (AA0XX0, 2H), 7.74 (d,
4
using
a
Micromass VG Quattro mass spectrometer with 3JH,H = 5.5 Hz, 1H), 7.32 (AA0XX0, 2H), 6.90 (d, JH,H = 2.2 Hz, 1H),
electrospray ionization (ESI), in series with a b1-flow detector. 6.71 (dd, JH,H = 2.2, JH,H = 8.7 Hz, 1H), 4.41–4.38 (m, 2H),
4
3
A
Jones Chromatography Genesiss C18, Scantec Lab, 4.24–4.21 (m, 2H), 2.73 (s, 3H), 2.33 (s, 3H). 13C NMR
10 cm ꢁ 4 mm, 4 mm HPLC column was used. The following (100 MHz, CD3OD 251c) d = 160.8, 146.4, 143.9, 142.1, 137.9,
mobile phases were used: 5 mM ammonium acetate (A) and 136.3, 134.4, 130.9, 130.1, 128.9, 123.5, 116.9, 113.3, 111.0, 96.9,
acetonitrile (B). Program: 10–90% (B) over 10 min, then 90% for 70.1, 67.2, 21.5, 19.4. ESI-MS: m/z = 397 [M1H]1.
5 min, flow rate 1 mL/min. The identities of compounds 5, 7, and
8
were confirmed by LC-MS analysis. Radioactivity was
3-[(1-Methyl-9H-b-carbolin-7-yl)oxy]propyl 4-methylbenzenesulfo-
nate (2)
measured in an ion chamber, Veenstra Instrumenten BV, VDC-
202. All chemicals were obtained from commercial suppliers and
used without further purification.
1-Methyl-9H-b-carbolin-7-ol (harmol) (0.050 g, 0.25 mmol) was
dissolved in 2 mL dry DMF. Anhydrous caesium carbonate
(0.124 g, 0.38 mmol) was added and the suspension was stirred
under a nitrogen atmosphere at room temperature for 30 min.
After cooling to -781C, a solution of propylene di(p-toluenesul-
fonate) (0.106 g, 0.28 mmol) in 2 mL dry DMF was added. The
reaction mixture was kept at -781C for 2 h before the
temperature was slowly raised to room temperature.
After another 16 h, the reaction mixture was extracted
with ethyl acetate (3 ꢁ 100 mL) and water (200 mL). The
combined organic phases were concentrated under reduced
pressure and the residue was purified by column chromato-
graphy (dichloromethane/methanol 10:1) yielding 2 as a yellow
oil (0.015 g, 15%). 1H NMR (400 MHz, CD3OD, 251C) d = 8.10
General labelling procedure
The precursor (1–4) (E15 mmol) was dissolved in 200 mL dry
DMF and added to a solution of the dried [K/K2.2.2.]118Fꢀ in
200 mL dry DMF. The reaction mixture was heated at 1501C for
15 min in a heating block, then diluted with 2 mL water and
purified by semi-preparative HPLC. Prior to autoradiography
experiments, the isolated compound was dissolved in 3 mL of a
solution containing phosphate buffer (pH 7.4)/PEG 4:1. In
metabolite experiments, only phosphate buffer (2 mL) was
used.
3
3
(d, JH,H = 5.5 Hz, 1H), 7.94 (d, JH,H = 8.7 Hz, 1H), 7.79
3
LC-MS data of labelled compounds
(d, JH,H = 5.5 Hz, 1H), 7.70 (AA0XX0, 2H), 7.18 (AA0XX0, 2H), 6.87
(5): Retention time = 7.1 min, m/z = 245 [M1H]1; (7): retention
time = 6.7 min, m/z = 289 [M1H]1; (8): retention time = 6.5 min,
m/z = 333 [M1H]1.
(d, 4JH,H = 2.0 Hz, 1H), 6.68 (dd, 4JH,H = 2.0, 3JH,H = 8.7 Hz, 1H), 4.27
(t, 3JH,H = 6.0 Hz, 2H), 3.99 (t, 3JH,H = 5.8 Hz, 2H), 2.77 (m, 3H), 2.15
(s, 3H), 2.14–2.09 (m, 2H). 13C NMR (100 MHz, CD3OD, 251C)
d = 161.2, 146.2, 143.9, 141.8, 137.5, 136.0, 134.0, 130.8, 128.6,
126.9, 123.3, 116.4, 113.2, 111.1, 96.3, 68.6, 64.6, 29.9, 21.4, 19.4.
ESI-MS: m/z = 411 [M1H]1.
Chemical synthesis
1H NMR, 13C NMR, 19F NMR, and HMBC spectra were recorded on
a Varian Unity (1H at 400 MHz, 13C at 100 MHz, 19F at 376 MHz) or
Varian Inova (1H at 500 MHz, 13C at 126 MHz) spectrometer using
CDCl3 or CD3OD as solvent (solvent peak used as reference,
except in the case of 19F where CFCl3 was used as reference). All
NMR experiments were conducted at 251C. Thin-layer chroma-
tography (TLC) was performed on Merck silica gel F-254
aluminium plates. Silica gel 60, particle size 0.040–0.063 mm
(Merck), was used for column chromatography. One millimetre
pre-coated plates of silica gel 60, F-254 (Merck), were used for
preparative TLC. LC-MS analyses were performed using a Gilson
HPLC and Finnigan AQA mass spectrometer in ESI mode.
7-[2-(2-Bromoethoxy)ethoxy]-1-methyl-9H-b-carboline (3)
1-Methyl-9H-b-carbolin-7-ol (harmol) (0.497 g, 2.51 mmol) was
dissolved in 10 mL dry DMF. Anhydrous caesium carbonate
(1.135 g, 3.79 mmol) was added and the suspension was stirred
under a nitrogen atmosphere at room temperature for 30 min.
After cooling to 01C, 1-bromo-2-(2-bromoethoxy)ethane (0.64 mL,
5.06 mmol) was added. The reaction mixture was kept at 01C for
1 h before the temperature was slowly raised to room temperature.
After another 3 h, the reaction mixture was extracted with ethyl
acetate (3 ꢁ 100 mL) and water (200 mL). The combined organic
phases were concentrated under reduced pressure and the residue
was purified by column chromatography (dichloromethane/
methanol 10:1.5) to yield 3 as a yellow oil (0.235 g, 25%). 1H
NMR (400 MHz, CDCl3, 251C) d= 9.60 (bs, 1H), 8.25 (d, 3JH,H =5.5Hz,
2-[(1-Methyl-9H-b-carbolin-7-yl)oxy]ethyl 4-methylbenzenesul-
fonate (1)
1-Methyl-9H-b-carbolin-7-ol (harmol) (1.497 g, 7.77 mmol) was
dissolved in 5 mL dry DMF. Anhydrous caesium carbonate
(3.721 g, 11.42 mmol) was added and the suspension was
stirred under a nitrogen atmosphere at room temperature for
1 h. After cooling to -781C, a solution of ethylene di(p-
toluenesulfonate) (3.093 g, 8.35 mmol) in 10 mL dry DMF was
added. The reaction mixture was kept at -781C for 2 h before
the temperature was slowly raised to room temperature. After
14 h more, the reaction mixture was extracted with ethyl
acetate (5 ꢁ 100 mL) and water (200 mL). The combined organic
phases were concentrated under reduced pressure and the
residue purified by column chromatography (dichloromethane/
3
3
1H), 7.93 (d, JH,H = 8.6 Hz, 1H), 7.72 (d, JH,H = 5.5 Hz, 1H), 7.07 (d,
4
4JH,H = 2.0 Hz, 1H), 6.90 (dd, JH,H =2.0, JH,H = 8.6 Hz, 1H), 4.18 (m,
2H), 3.88 (m, 4H), 3.48 (t, JH,H =6.4Hz, 2H), 2.86 (s, 3H). 13C NMR
3
3
(100 MHz, CDCl3, 251C) d = 159.6, 142.5, 140.8, 136.5, 134.8, 128.5,
122.3, 115.1, 112.1, 110.1, 95.4, 71.0, 69.2, 67.4, 30.3, 19.9. ESI-MS:
m/z (%) = 349 (50.5), 351 (49.5) [M1H]1.
7-{2-[2-(2-Chloroethoxy)ethoxy]ethoxy}-1-methyl-9H-b-carboline
(4)
1-Methyl-9H-b-carbolin-7-ol (harmol) (0.500 g, 2.52 mmol) was
dissolved in 10 mL dry DMF. Anhydrous caesium carbonate
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 277–282

E. Blom et al.
(0.985 g, 3.02 mmol) was added and the suspension was stirred 7-[2-(2-Fluoroethoxy)ethoxy]-1-methyl-9H-b-carboline (unlabelled
reference to 7)
under a nitrogen atmosphere at room temperature for 30 min.
After cooling to 01C, 1-chloro-2-[2-(2-chloroethoxy)ethox-
y]ethane (0.943 g, 5.04 mmol) was added. The reaction mixture
was kept at 01C for 2 h before the temperature was slowly raised
to room temperature. After another 24 h, the reaction mixture
was extracted with ethyl acetate (3 ꢁ 100 mL) and water
(200 mL). The combined organic phases were concentrated
under reduced pressure and the residue was purified by
column chromatography (dichloromethane/methanol 10:1.5)
to yield 4 as a colourless oil (0.037 g, 4%). 1H NMR (400 MHz,
7-[2-(2-Bromoethoxy)ethoxy]-1-methyl-9H-b-carboline
(3)
(0.085 g, 0.29 mmol) was dissolved in 5 mL dry THF. Dry DMF
(0.1 mL) and tetrabutylammonium fluoride (0.6 mL, 1 M in
THF) were added. The reaction mixture was heated at 651C for
8.5 h and then extracted with ethyl acetate (3 ꢁ 50 mL) and
water (100 mL). The combined organic phases were concen-
trated under reduced pressure and the residue was purified
by preparative TLC (dichloromethane/methanol 10:2) to yield
7-[2-(2-fluoroethoxy)ethoxy]-1-methyl-9H-b-carboline as
a
3
CDCl3, 251C) d = 9.21 (bs, 1H), 8.29 (d, JH,H = 5.6 Hz, 1H), 7.93 (d,
colourless oil (0.004 g, 5 %). 1H NMR (400 MHz, CDCl3, 251C)
3
3JH,H = 8.7 Hz, 1H), 7.72 (d, JH,H = 5.6 Hz, 1H), 7.00 (4JH,H = 2.2 Hz,
3
3
d = 8.31 (d, JH,H = 5.2 Hz, 1H), 7.97 (d, JH,H = 8.8 Hz, 1H), 7.73
(d, JH,H = 5.2 Hz, 1H), 7.02 (d, JH,H = 2.2 Hz, 1H), 6.94
4
3
1H), 6.90 (dd, JH,H = 2.2, JH,H = 8.7 Hz, 1H), 4.21–4.18 (m, 2H),
3
4
3
3.91–3.88 (m, 2H), 3.77–3.69 (m, 6H), 3.61 (t, JH,H = 6.1 Hz, 2H),
4
3
2
(dd, JH,H = 2.2, JH,H = 8.8 Hz, 1H), 4.62 (dm, JH,F = 47.6 Hz,
2H), 4.30–4.27 (m, 2H), 4.00–3.97 (m, 2H), 3.86–3.84 (m, 2H),
2.83 (s, 3H). 13C NMR (100 MHz, CDCl3, 251C) d = 159.9, 141.7,
140.9, 138.5, 134.7, 128.6, 122.6, 116.0, 112.2, 110.0, 95.7, 84.0,
82.3, 70.7, 70.5, 67.8, 66.3, 20.0 ppm. 19F NMR (376 MHz,
CDCl3, 251C) d = ꢀ222.0 (m). ESI-MS: m/z = 289 [M1H]1.
2.83 (s, 3H). 13C NMR (100 MHz, CDCl3, 251C) d = 160.2, 142.2,
140.7, 137.7, 134.8, 129.0, 122.6, 115.9, 112.3, 110.5, 95.9, 71.4,
70.9, 70.8, 69.8, 68.0, 42.7, 19.8. ESI-MS: m/z (%) = 348 (75.5), 350
(24.5) [M1H]1.
7-(2-Fluoroethoxy)-1-methyl-9H-b-carboline (unlabelled reference
to 5)
7-{2-[2-(2-Fluoroethoxy)ethoxy]ethoxy}-1-methyl-9H-b-carboline
(unlabelled reference to 8)
2-[(1-Methyl-9H-b-carbolin-7-yl)oxy]ethyl 4-methylbenzenesulfo-
nate (1) (0.020g, 0.050 mmol) was dissolved in 0.5 mL dry THF
and tetrabutylammonium fluoride (0.03 mL, 1 M in THF) was
added. The reaction mixture was heated at 651C for 2 h and then
extracted with ethyl acetate (3ꢁ 50mL) and water (100mL). The
combined organic phases were concentrated under reduced
pressure and the residue was purified by preparative TLC
(dichloromethane/methanol 10:1) to yield 7-(2-fluoroethoxy)-1-
7-{2-[2-(2-Chloroethoxy)ethoxy]ethoxy}-1-methyl-9H-b-carbo-
line (4) (0.020 g, 0.058 mmol) was dissolved in 2 mL dry THF
and tetrabutylammonium fluoride (0.5 mL, 1 M in THF) was
added. The reaction mixture was heated at 651C for 48 h and
then extracted with ethyl acetate (3 ꢁ 50 mL) and water
(100 mL). The combined organic phases were concentrated
under reduced pressure and the residue was purified by
preparative TLC (dichloromethane/methanol 10:1) to yield 7-
{2-[2-(2-fluoroethoxy)ethoxy]ethoxy}-1-methyl-9H-b-carboline
as a colourless oil (0.002 g, 11%). 1H NMR (400 MHz, CDCl3,
1
methyl-9H-b-carboline as a colourless oil (0.002g, 16%). H NMR
(500MHz, CDCl3, 251C) d = 8.34 (d, 3JH,H = 5.3 Hz, 1H), 8.22 (bs, 1H),
7.98 (d, JH,H = 8.9 Hz, 1H), 7.72 (d, JH,H = 5.3 Hz, 1H), 6.99 (d,
3
3
4JH,H = 2.1 Hz, 1H), 6.93 (dd, 4JH,H = 2.1, 3JH,H = 8.9 Hz, 1H), 4.83 (dm,
3
3
251C) d = 8.16 (d, JH,H = 5.0 Hz, 1H), 7.84 (d, JH,H = 8.8 Hz, 1H),
7.63 (d, JH,H = 5.0 Hz, 1H), 7.23 (d, JH,H = 2.1 Hz, 1H), 6.78
3
2JH,F = 47.3Hz, 2H), 4.33 (dm, JH,F = 27.6Hz, 2H), 2.80 (s, 3H). 13C
3
3
NMR (100MHz, CDCl3, 251C) d = 159.5, 141.4, 141.1, 139.0, 134.7,
128.5, 122.8, 116.4, 112.3, 109.8, 95.9, 81.9 (d, 1JC,F = 172.3 Hz), 67.5
3
2
(dm, JH,H = 8.8 Hz, 1H), 4.50 (dm, JH,F = 47.7 Hz, 2H), 3.78–3.55
(m, 8H), 3.20–3.13 (m, 2H), 2.91 (s, 3H). 13C NMR (100 MHz,
CDCl3, 251C) d = 159.6, 142.8, 141.5, 136.6, 135.0, 128.2, 121.9,
115.1, 111.7, 110.1, 96.0, 83.9, 82.2, 71.2, 70.6, 70.4, 70.2, 69.6,
67.6, 20.5. 19F NMR (376 MHz, CDCl3, 251C) d = ꢀ222.9 (m). ESI-
MS: m/z = 333 [M1H]1.
2
(d, JC,F = 20.7 Hz), 20.2. 19F NMR (376MHz, CDCl3, 251C)
d = ꢀ222.3 (m). ESI-MS: m/z = 245 [M1H]1.
7-(2-Fluoropropoxy)-1-methyl-9H-b-carboline (unlabelled reference
to 6)
3-[(1-Methyl-9H-b-carbolin-7-yl)oxy]propyl 4-methylbenzene-
sulfonate (2) (0.517 g, 1.26 mmol) was dissolved in 5 mL dry
DMF and tetrabutylammonium fluoride (1.26 mL, 1 M in THF)
was added. The reaction mixture was heated at 901C for 2.5 h
and then extracted with ethyl acetate (3 ꢁ 50 mL) and water
(100 mL). The combined organic phases were concentrated
under reduced pressure and the residue was purified by
column chromatography (dichloromethane/methanol 10:1) to
yield 7-(2-fluoropropoxy)-1-methyl-9H-b-carboline as a colour-
less oil (0.070 g, 22%).1H NMR (400 MHz, CDCl3, 251C) d = 8.27
Biology
Autoradiography
Brains from male Sprague–Dawley rats were sectioned at 25 mm
using a microtome (Microm HM 560, Microm, Germany) and
stored at -201C until used. Sections with striatum were
incubated in 0.01–5 nM tracer in phosphate-buffered saline
(PBS) buffer for 40 min at room temperature. For the determina-
tion of non-specific binding, 1 mM harmine (ꢂ500 ꢁ kd for
[11C]HAR, 2.070.7 nM14) was added to the incubate. The
sections were washed 3 times in PBS buffer on ice for 3 min,
briefly washed in distilled water, dried for approximately 10 min
at 371C, and placed on a phosphor imager plate (Amersham
Biosciences, USA) for exposure for 2 h (approximately one half-
life). The plates were scanned at 100 mm resolution using a
Phospho imager Model 400S (Molecular Dynamics, USA). The
autoradiograms were analyzed using ImageQuant 5.1 (Molecu-
3
3
(d, JH,H = 5.4 Hz, 1H), 7.93 (d, JH,H = 8.7 Hz, 1H), 7.71 (d,
4
3JH,H = 5.4 Hz, 1H), 6.95 (d, JH,H = 2.2 Hz, 1H), 6.87 (dd,
3
4JH,H = 2.2, JH,H = 8.7 Hz, 1H), 4.63 (dt, 3JH,H = 5.9, 2JH,F = 47.0 Hz,
3
2H), 4.11 (t, JH,H = 6.2 Hz, 2H), 2.80 (s, 3H), 2.16 (dm,
3JH,F = 25.6 Hz, 2H). 13C NMR (100 MHz, CDCl3, 251C) d = 160.1,
142.2, 141.0, 138.0, 135.0, 128.9, 122.6, 115.9, 112.3, 110.1, 95.7,
1
3
80.7 (d, JC,F = 164.6 Hz), 64.0 (d, JC,F = 5.3 Hz), 30.4 (d,
2JC,F = 20.1 Hz), 19.8. 19F NMR (376 MHz, CDCl3, 251C)
d = ꢀ222.5 (m). ESI-MS: m/z = 259 [M1H]1.
J. Label Compd. Radiopharm 2008, 51 277–282
Copyright r 2008 John Wiley & Sons, Ltd.
www.jlcr.org

E. Blom et al.
¨
lar Dynamics, USA). Aliquots (20 mL) of the incubation solution Bergstrom-Pettermann is acknowledged for performing a part of
were dropped onto a filter paper and scanned together with the the biological experiments.
sections to be used as a calibration. Two to five experiments,
each with sections from one or two different rats, were
performed for each tracer. Four concentrations were used in
each experiment.
References
Metabolite study
[1] R. H. F. Manske, The carboline alkaloids in The Alkaloids, Vol. VIII,
Academic Press, New York, 1965, pp. 47–53 (Chapter 3).
[2] T. J. Hagen, P. Skolnick, J. M. Cook, J. Med. Chem. 1987, 30,
750–753.
[3] Y. Song, J. Wang, S. F. Teng, D. Kesuma, Y. Deng, J. Duan, J. H.
Wang, R. Zhong-Qi, M. M. Sim, Bioorg. Med. Chem. Lett. 2002, 12,
1129–1132.
[4] H. Kim, S. O. Sablin, R. R. Ramsay, Arch. Biochem. Biophys. 1997,
337, 137–142.
[5] H. Blaschko, Rev. Physiol. Biochem. Pharmacol. 1974, 70,
83–148.
[6] J. A. Fuentes, N. H. Neff, Neuropharmacology 1975, 14, 819–825.
After the radioactive compound in the buffer solution was
injected into a Sprague–Dawley rat, blood samples were
withdrawn at predefined times and centrifuged. After addition
of acetonitrile containing isotopically unmodified 7 or 8,
respectively, to the plasma samples, centrifugation, and filtering,
the samples were analyzed using a Beckman Coulter Ultraphere
ODS 250 ꢁ 10 mm, 5 mm LC column at a flow of 5 mL/min.
Acetonitrile (20%) and 0.05 M ammonium formate pH 3.5 (80%)
were used as eluent.
˚
[7] M. Bergstrom, G. Westerberg, T. Kihlberg, B. Langstrom, Nucl. Med.
¨
Biol. 1997, 24, 381–388.
¨
¨
´
[8] M. Bergstrom, G. Westerberg, G. Nemeth, M. Traut, G. Gross, G.
Greger, H. Mu¨ller-Peltzer, A. Safer, S.-A. Eckernas, A. Grahner, B.
Conclusion
˚
¨
´
˚
¨
Langstrom, Eur. J. Clin. Pharmacol. 1997, 52, 121–128.
[9] N. Ginovart, J. H. Meyer, A. Boovariwala, D. Hussey, E. A. Rabiner,
S. Houle, A. A. Wilson, J. Cereb. Blood Flow Metab. 2006, 26,
330–344.
Four 18F-labelled harmine analogues have been synthesized
using a one-step nucleophilic 18F-fluorination strategy. Pure
labelled compound was obtained in less than 70 min from the
end of radionuclide production, with a decay-corrected yield up
to 23%. Specific radioactivity was in the range 400–700 GBq/
mmol. The two analogues 7 and 8, having PEG side chains,
showed higher specific binding to MAO-A in in vitro experiments
than did the corresponding tracer [11C]HAR.14 The fraction of
unmetabolized labelled tracer in rat blood plasma was highest
for compound 8, when compared with the other pegylated
analogue 7, and will be used in further studies.
[10] S. B. Jensen, A. K. Olsen, K. Pedersen, P. Cumming, Synapse 2006,
59, 427–434.
[11] R. Murthy, K. Erlandsson, D. Kumar, R. Van Heertum, J. Mann,
R. Parsey, Nucl. Med. Commun. 2007, 28, 748–754.
¨
[12] H. Orlefors, A. Sundin, K.-J. Fasth, K. Oberg, B. Langstrom,
¨
˚
B. Eriksson, M. Bergstrom, Nucl. Med. Biol. 2003, 30, 669–679.
¨
¨
[13] W. Zhang, S. Oya, M.-P. Kung, C. Hou, D. L. Maier, H. F. Kung, Nucl.
Med. Biol. 2005, 32, 799–809.
˚
[14] M. Bergstrom, G. Westerberg, B. Langstrom, Nucl. Med. Biol. 1997,
¨
24, 287–293.
¨
[15] J. Mukherjee, Z.-Y. Yang, Nucl. Med. Biol. 1999, 26, 619–625.
[16] D. W. Kim, D.-S. Ahn, Y.-H. Oh, S. Lee, H. S. Kil, S. J. Oh, S. J. Lee, J. S. Kim,
J. S. Ryu, D. H. Moon, D. Y. Chi, J. Am. Chem. Soc. 2006, 128,
16394–16397.
Acknowledgement
[17] www.acdlabs.com.
[18] P. Bjurling, R. Reineck, G. Westerberg, A. Gee, J. Sutcliffe, B.
Financial support from The Swedish Science Research Council
and National Institutes of Health is gratefully acknowledged.
This work was conducted in collaboration with Uppsala Imanet
and Uppsala Applied Science Lab, GE Healthcare. Elisabeth
˚
¨
Langstrom, Presented at Proceedings of the VIth Workshop on
Targetry and Target Chemistry, TRIUMF, Vancouver, Canada, 1995,
pp. 282–284.
www.jlcr.org
Copyright r 2008 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2008, 51 277–282