Synthesis of tritium-labeled levetiracetam ((2S)-2-(2-oxopyrrolidin-1-yl) butanamide) with high specific activity

Article abstract of DOI:10.1002/jlcr.1942

A method for the preparation of [A³H]levetiracetam with a high specific activity of 98Ci/mmol (3.6TBq/mmol) is described. The radioligand proved to be highly useful for the labeling of specific levetiracetam binding sites in rat brain membrane preparations.

Full text of DOI:10.1002/jlcr.1942

Note
Received 16 June 2011,
Revised 30 August 2011,
Accepted 17 September 2011
Published online 18 October 2011 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.1942
Synthesis of tritium-labeled levetiracetam
((2S)-2-(2-oxopyrrolidin-1-yl)butanamide) with
high specific activity
*
Simone Hildenbrand, Younis Baqi, and Christa E. Müller
A method for the preparation of [³H]levetiracetam with a high specific activity of 98 Ci/mmol (3.6 TBq/mmol) is described.
The radioligand proved to be highly useful for the labeling of specific levetiracetam binding sites in rat brain membrane
preparations.
Keywords: levetiracetam; tritium labeling; high specific activity; SV2A protein; epilepsy
provide an improved tool for studying the interaction of LEV
with the SV2A protein, including SV2A mutants, as well as further
potential targets of the antiepileptic drug LEV (Figure 1).
Introduction
Levetiracetam (1; (2S)-2-(2-oxopyrrolidin-1-yl)butanamide, KeppraW)
is a pyrrolidone drug widely used in antiepileptic therapy. Since
its approval in the USA in 1999, it has become one of the most
successful therapeutics of the newer classes of antiepileptic
Results and discussion
drugs (AED). Several beneficial properties including potent
antiepileptogenic effects, absence of severe adverse effects,
and a low potential for drug interactions1,2 have indicated that
levetiracetam (LEV) acts via a novel, unique mechanism of action,
which is distinct from that of conventional AEDs. In 2004, it was
discovered that LEV binds to the synaptic vesicle protein SV2A.3
However, until today it is still unknown how the interaction of
LEV with the SV2A protein is translated into its antiepileptic
properties. Furthermore, interaction with other yet unknown
targets may contribute to LEV’s pharmacologic effects.
To study the interaction between drugs and their protein
targets, radioligands have been highly useful. LEV has previously
been prepared in ³H-labeled form ([³H]LEV) with a specific
activity of 36.6 Ci/mmol (1.35 TBq/mmol), which was used for
studying LEV binding to SV2A protein in native rat tissues.4 This
radioligand, however, was not ideal for our planned biological
investigations because of its relatively low affinity for the SV2A
protein (~1 mM) combined with its moderate specific activity.
Since the SV2A protein has a very high abundance (expression
levels of, e.g., ~9 pmol/mg of protein in rat brain membrane
preparations),4 further potential target proteins (e.g. membrane
receptors) that typically show at least 10-fold lower expression
levels can hardly be detected with this radioligand. Derivatives
of LEV, [³H]ucb 30889 ((2S)-2-[4-(3-azidophenyl)-2-oxopyrrolidin-
1-yl]butanamide, KD = 62 nM, 32 Ci/mmol, 1.2 TBq/mmol)5 and, very
recently, [3H]brivaracetam (2S-2-[(4R)-2-oxo-4-propylpyrrolidin-1-yl]
butanamide, KD = 80 nM, 8 Ci/mmol, 0.3 TBq/mmol),6 have also
been prepared as tritium-labeled radioligands.
The synthesis of [³H]LEV (7) with high specific activity was per-
formed as depicted in Scheme 1. A suitable LEV precursor (com-
pound 6; see Scheme 1) was prepared for radiolabeling, which
would allow for the introduction of four tritium atoms by cata-
lytic hydrogenation. Thus, (S)-2-aminobutyric acid 2 was reacted
with thionyl chloride in methanol to yield the methyl ester 3 in
nearly quantitative yield.7 The methyl ester was then transformed
to the corresponding amide 4 in a microwave reaction using
ammonia in methanol. Condensation of 4 with mucochloric
acid 5 in the presence of sodium triacetoxyborohydride and
catalytic amounts of acetic acid in chloroform yielded lactam
6,8 which served as precursor molecule for the generation of
the radioligand. Compound 6 was custom-labeled by Quotient
3
Bioresearch, UK, by catalytic hydrogenation with H2 gas in the
presence of palladium on charcoal as a catalyst. Owing to the
exchange of the two chlorine atoms as well as a catalytic reduc-
tion of the double bond, it was possible to introduce a maximum
of four tritium atoms per molecule and to obtain [³H]LEV with a
specific activity of 98 Ci/mmol (3.6 TBq/mmol). Thus, the above-
mentioned precursor molecule 6 allows the preparation of a
radioligand 7 with very high specific activity, almost 3-fold
higher than that of the previously published tritiated LEV.4
The binding properties of the radioligand were investigated in
radioligand binding assays using rat brain cortical membrane
PharmaCenter Bonn, Pharmaceutical Institute, Pharmaceutical Chemistry I,
Pharmaceutical Sciences Bonn (PSB), University of Bonn, An der Immenburg 4,
D-53121 Bonn, Germany
In the present study, we aimed at obtaining a tritiated form of
the clinically applied drug of this important chemical class of
AEDs, [3H]LEV, with a significantly higher specific activity as com-
pared with the previously prepared radioligand. Our goal was to
*Correspondence to: Dr. Christa E. Müller, Pharmazeutisches Institut,
Pharmazeutische Chemie I, An der Immenburg 4, D-53121 Bonn, Germany.
E-mail: christa.mueller@uni-bonn.de
J. Label Compd. Radiopharm 2012, 55 48–51
Copyright © 2011 John Wiley & Sons, Ltd.

S. Hildenbrand et al.
Figure 1. Structure of levetiracetam.
Figure 3. Specific binding of [3H]levetiracetam obtained in saturation binding
experiments using rat brain cortical membrane preparations. Incubation was per-
formed for 120 min at 4ꢀC using 200 mg of protein per vial. For this experiment,
the radioligand was diluted with unlabeled levetiracetam (isotopic dilution). The
graph shows the mean curve of two individual experiments performed in tripli-
cates. A KD value of 1.12 Æ 0.12 mM and a Bmax value of 3.8 Æ 0.1 pmol/mg protein
were calculated.
Scheme 1. Synthesis of [3H]levetiracetam.
preparations (Figures 2–4). Nonspecific binding was determined
in the presence of a high concentration of unlabeled levetirace-
tam (1 mM). The obtained results showed that the binding was
concentration as well as protein dependent. A protein concentra-
tion of 200 mg/assay and a radioligand concentration of 10 nM
yielded the best results, affording a specific binding of ~85%
(Figure 2). In saturation binding experiments, the radioligand
showed saturable binding, with a KD value of 1.12 Æ 0.12 mM
and a Bmax value of 3.8 Æ 0.1 pmol/mg protein (Figures 3 and 4),
compatible with previously published results.4
Figure 4. Total, nonspecific, and specific binding (cpm) of [3H]levetiracetam
obtained in saturation binding experiments using rat brain cortical membrane
preparations. Incubation conditions were as described in Figure 3. The graphs
show the mean curves of two individual experiments performed in triplicates.
Experimental
General
All solvents and reagents were obtained from commercial
sources and not further purified unless indicated otherwise.
The microwave reaction was performed in a Discover microwave
(CEM GmbH, Kamp-Lintfort).
Analysis of unlabeled compounds. Analytical thin layer chroma-
tography (TLC) was performed on silica-coated aluminum plates
containing a fluorescent indicator (Merck silica gel 60 F254), and
detection was achieved with UV light (254 nm) and by spraying
with Ninhydrin9 reagent. Low-resolution mass spectra were
obtained on an API 2000 mass spectrometer (electrospray ioniza-
tion, Applied Biosystems, Darmstadt, Germany) coupled to an
HPLC system (Agilent 1100) using the following procedure:
Compounds were dissolved in methanol (0.5 mg/ml), and a
10-ml sample of this solution was injected into the HPLC system
containing a Phenomenex Luna C18 column (50 mm  2.00 mm,
particle size 3 mm). It was chromatographed using a gradient of
water/methanol from 90:10 to 0:100 in 30 min. The gradient
was started after 10 min; the flow rate was 250 ml/min. UV
absorption was detected using a diode array detector (from
190 to 900 nm), and purity was determined at 254 nm. NMR
Figure 2. Nonspecific and specific binding (cpm) of [3H]levetiracetam in rat brain
cortical membrane preparations: 100, 200, or 300 mg of protein was incubated for
120 min at 4ꢀC with [3H]levetiracetam (5 and 10 nM). Nonspecific binding (open
bars) was determined in the presence of unlabeled levetiracetam (1 mM). Specific
binding (dotted bars) was obtained by subtraction of nonspecific binding from
total binding, which was determined in the absence of unlabeled levetiracetam.
J. Label Compd. Radiopharm 2012, 55 48–51
Copyright © 2011 John Wiley & Sons, Ltd.
www.jlcr.org

S. Hildenbrand et al.
spectra were recorded on a Bruker Avance 500 spectrometer (CON), 171.40 (CONH2). LC-MS m/z 237 ([M + H]+), [a]D20 = À21.2
(1H: 500 MHz, 13C: 125 MHz) at room temperature. Spectra were (c = 0.33, CHCl3).
recorded in CDCl3, and the remaining protons of the deuterated
(S)-[3,3,4,4-³H]2-(2-Oxopyrrolidin-1-yl)butanamide (7)
solvent were used as an internal standard (1H: d (ppm) CDCl3:
7.24 and 13C: d (ppm) CDCl3: 77.0). Coupling constants are given
The labeling of 6 with tritium was performed by Quotient Biore-
search using the following procedure: Compound 6 (3 mg) and
10% palladium on charcoal (20 mg) were stirred in ethanol
(2 ml) containing N,N-diisopropylethylamine (100 ml) in the pre-
sence of tritium gas (10 Ci) for 4 h. Labile tritium was removed
by repeated evaporation to dryness with ethanol. The crude
yield was 850 mCi, and the radiochemical purity was 60%. Purifi-
cation of the radioligand was performed by HPLC (detection at
205 nm), giving a radiochemical purity of 99.7%. The specific
activity was determined to be 98 Ci/mmol (3.6 TBq/mmol).
in hertz (Hz), and chemical shifts in parts per million (ppm). Spin
multiplicities are abbreviated as s (singlet), d (doublet), t (triplet),
q (quartet), m (multiplet), and br (broad).
Analysis of the radioligand 7. Analysis was performed by
Quotient Bioresearch, UK, by HPLC, using a Luna C18 column
(150 Â 4.6 mm, particle size 3 mm) applying a gradient of water/
methanol from 90:10 to 0:100 in the presence of ammonium
acetate (2 mM) within 20 min with a flow rate of 500 ml/min,
and by mass spectrometry. Chemical purity was detected at
205 nm.
Radioligand binding studies
Chemistry
Radioligand binding studies were performed in analogy to the
procedure described by Noyer et al.4 Rat brain cortical mem-
brane preparations (100–300 mg/vial) were incubated for
120 min at 4 ꢀC in a total volume of 500 ml of Tris-HCl buffer solu-
tion (50 mM, pH 7.4), containing 2 mM MgCl2 and [³H]LEV (5 and
10 nM, respectively). Nonspecific binding was determined in the
presence of unlabeled levetiracetam (1 mM). For saturation
experiments, 200 mg of rat brain cortical membrane preparations
was incubated with increasing amounts of radioligand, which
was diluted with unlabeled levetiracetam (isotopic dilution).
Separation of bound from unbound radioligand was achieved
by filtration through GF/C glass fiber filters pre-soaked for
30 min in 0.1% aqueous polyethyleneimine solution. The filters
were subsequently dried (for 90 min at 50 ꢀC), and the remaining
radioactivity was determined by liquid scintillation counting at a
counter efficiency of 0.55. Data were analyzed with GraphPad
PrismW 5.01 (GraphPad Software, San Diego, CA, USA).
(S)-Methyl 2-aminobutanoate hydrochloride (3)
Freshly distilled thionyl chloride (30 mmol) was added dropwise
to 10 ml of methanol previously cooled to À20 ꢀC. After addition
of 2 (10 mmol), the mixture was stirred at room temperature. The
reaction progress was monitored by TLC, and if needed,
additional thionyl chloride was added until the reaction was
1
completed, giving an almost quantitative yield. H NMR d ppm
1.08 (t, 3H, J = 7.43, CH3), 2.07–2.13 (m, 2H, CH2), 3.78 (s, 3H,
OCH3), 4.09–4.13 (m, 1H, CH), 8.73 (s, 3H, NH+3ClÀ). 13C NMR d
ppm 9.63 (CH3), 23.78 (CH2), 53.04 (OCH3), 54.40 (CH), 169.77
(C=O).
(S)-2-Aminobutanamide (4)
A sample of 3 (1 mmol) was dissolved in 6 ml of ammonia (7 M)
in methanol. The solution was stirred under microwave irradia-
tion (70 W) at 100 ꢀC for 240 min. The mixture was evaporated
to dryness, and the residue was purified by column chromato-
graphy (gradient of dichloromethane/methanol from 100:0 to
70:30, containing 2% aqueous ammonia solution). Pure fractions
were combined, evaporated to dryness, and subsequently
Conclusions
A synthetic route has been devised that allows for the prepara-
tion of ³H-labeled levetiracetam with a high specific activity of
98 Ci/mmol (3.6 TBq/mmol), which is nearly 3-fold higher than
that of the previously prepared tritiated LEV. The radioligand
has been successfully used in initial radioligand binding
studies with rat brain cortical membrane preparations and
was shown to be a useful tool for the specific labeling of high-
affinity binding sites for levetiracetam, which has previously
been identified as the vesicular protein SV2A. The described
radioligand should be applicable for further investigations
on native tissues including human brain samples as well as
recombinantly expressed SV2A protein. Moreover, besides
determining binding to the SV2A protein, this new radioligand
should also be suitable for the detection of potential target
proteins of much lower abundance than SV2A, and may thus
contribute to getting more insight into the mechanisms of
action of LEV.
1
dissolved in water and lyophilized (yield 77%). H NMR d ppm
0.95 (t, 3H, J = 7.55, CH3), 1.52–1.60 (m, 1H + 2H, CH2 + NH2),
1.78–1.87 (m, 1H, CH2), 3.28–3.30 (m, 1H, CH), 5.79, 7.03 (2 s, 1H
each, CONH2). 13C NMR d ppm 9.94 (CH3), 27.97 (CH2), 56.32 (CH),
178.16 (C=O). Obtained data corresponded to published data.10
(S)-2-(3,4-Dichloro-2,5-dihydro-2-oxo-1H-pyrrol-1-yl)butanamide (6)
Mucochloric acid 5 (2 mmol) and compound 4 (2 mmol) were dis-
solved in a mixture of 10 ml of chloroform and 0.2 ml of acetic
acid. After addition of sodium triacetoxyborohydride (3 mmol,
1.5 equiv.), the mixture was stirred at room temperature for
several hours. The reaction progress was monitored by TLC. After
approximately 20 h, a saturated solution of ammonium chloride
(20 ml) was added to the reaction mixture. The product was
extracted from the aqueous phase with chloroform (3 Â 20 ml).
The organic layers were collected and washed with water
(20 ml) and subsequently with brine (10 ml). After drying over
sodium sulfate, the solvent was evaporated, and the residue pur-
ified by column chromatography eluting with cyclohexane/ethyl
acetate (1:4). Yield 75% (lit. yield 62%).8 1H NMR d ppm 0.94 (t, 3H,
J = 7.43, CH3), 1.70–1.79, 1.95–2.04 (2 m, each 1H, CH2), 4.04, 4.34
(AB-system, 2H, J = 18.9, CH2-N), 4.55–4.59 (m, 1H, CH), 5.42, 6.17
(2br, each 1H, NH2). 13C NMR d ppm 10.42 (CH3), 22.42 (CH2),
Acknowledgement
The authors would like to thank the International Isotope
Society-Central European Division (IIS-CES) and the Deutsche
Forschungsgemeinschaft (Graduiertenkolleg 804) for financial
support. We would like to thank Nicole Florin for her expert
50.98 (CH2-N), 56.39 (CH), 124.73 (CCl-CO), 141.37 (CH2-CCl), 165.00 assistance in the radioligand binding studies.
www.jlcr.org
Copyright © 2011 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 48–51

S. Hildenbrand et al.
[5] M. Gillard, B. Fuks, P. Michel, P. Vertongen, R. Massingham,
P. Chatelain, Eur. J. Pharmacol. 2003, 478, 1–9.
[6] M. Gillard, B. Fuks, K. Leclercq, A. Matagne, Eur. J. Pharmacol. 2011,
664, 36–44.
References
[1] W. Löscher, D. Hönack, C. Rundfeldt, J. Pharmacol. Exp. Ther. 1998,
284, 474–479.
[2] S. Shorvon, Lancet 2001, 358, 1885–1892.
[3] B. A. Lynch, N. Lambeng, K. Nocka, P. Kensel-Hammes, S. M. Bajja-
lieh, A. Matagne, B. Fuks, Proc. Natl. Acad. Sci. USA 2004, 101,
9861–9866.
[7] M. E. F. Braibante, H. T. S. Braibante, A. F. Morel, C. C. Costa, M. G.
Lima, J. Braz. Chem. Soc. 2006, 17, 184–188.
[8] K. Das Sarma, J. Zhang, Y. Huang, J. G. Davidson, Eur. J. Org. Chem.
2006, 37, 3730–3737.
[4] M. Noyer, M. Gillard, A. Matagne, J.-P. Hénichart, E. Wülfert, Eur. J.
Pharmacol. 1995, 286, 137–146.
[9] V. J. Harding, R. M. MacLean, J. Biol. Chem. 1916, 25, 337–350.
[10] P. V. R. Acharyulu, C. M. H. Raju, PCT Int. Appl. US 2005/0182262 A1, 2005.
J. Label Compd. Radiopharm 2012, 55 48–51
Copyright © 2011 John Wiley & Sons, Ltd.
www.jlcr.org