Development of SPECT imaging agents for the norepinephrine transporters: [123I]INER

Article abstract of DOI:10.1016/j.bmcl.2006.10.018

A series of reboxetine analogs was synthesized and evaluated for in vitro binding as racemic mixtures. The best candidate (INER) was synthesized as the optically pure (S,S) enantiomer, labeled with iodine-123 and its in vivo binding determined by SPECT im

Full text of DOI:10.1016/j.bmcl.2006.10.018

Bioorganic & Medicinal Chemistry Letters 17 (2007) 533–537
Development of SPECT imaging agents
for the norepinephrine transporters: [ I]INER
1
23
q
a,b,
b
a
b
*
Gilles D. Tamagnan,
Eric Brenner, David Alagille, Julie K. Staley,
b
a
a
b
c
Colin Haile, Andrei Koren, Michelle Early, Ronald M. Baldwin, Frank I. Tarazi,
c
d
d
a,b
Ross J. Baldessarini, Nachwa Jarkas, Mark M. Goodman and John P. Seibyl
a
Institute for Degenerative Diseases, New Haven, CT 06520, USA
Yale University School of Medicine, VA Connecticut HCS (116A2), 950 Campbell Avenue, CT 06516, USA
b
c
Department of Psychiatry and Neuroscience Program, Harvard Medical School and Mailman Research Center,
McLean Division of Massachusetts General Hospital, Belmont, MA 02478-9106, USA
d
Emory University, Atlanta, GA, USA
Received 25 August 2006; revised 4 October 2006; accepted 5 October 2006
Available online 12 October 2006
Abstract—A series of reboxetine analogs was synthesized and evaluated for in vitro binding as racemic mixtures. The best candidate
INER) was synthesized as the optically pure (S,S) enantiomer, labeled with iodine-123 and its in vivo binding determined by
(
1
23
SPECT imaging in baboons. The in vivo specificity, selectivity, and kinetics of [ I]INER make it a promising agent for imaging
NET in vivo by noninvasive SPECT imaging.
Ó 2006 Elsevier Ltd. All rights reserved.
The norepinephrine transporter (NET) is a protein with
2 transmembrane-spanning domains, located at the
development of a specific radioligand to quantify the
variation of NET density in vivo will be of great utility
as a diagnostic tool to better understand the etiology
and pathogenesis of these neuropsychiatric disorders,
as well as a tool to evaluate potential new drugs target-
ing the NET.
1
pre-synaptic terminal of noradrenergic neurons. The
principal physiological function of NET is to remove ex-
cess norepinephrine (NE) from the synaptic cleft to ter-
minate the action of NE (avoiding over-stimulation) and
to recycle NE into the pre-synaptic neuron (active reup-
take) for later re-release. Brain structures known to be
rich in NET include the locus coerulus (brainstem area);
thalamus, hippocampus, and throughout the cerebral
cortex, whereas low densities are found in cerebellum
Research on quantitative mapping of dopamine trans-
porter (DAT) and serotonin transporter (SERT) related
to various CNS disorders has benefited from the avail-
ability of suitable radioligands (e.g., b-CIT as DAT
4
13,14
and striatum. NET has been implicated in the patho-
radioligand,
1
and DASB or ZIENT as SERT radio-
). In contrast, brain imaging of NET has
5,16
physiology of numerous neuropsychiatric and neurode-
generative disorders, including depression (reduced
level of NE in the synapse appears to down-regulate
ligands
been hampered by lack of a suitable radioligand. How-
ever, in the past few years some selective NET radioli-
gands have been prepared and evaluated in animals.
5
,6
7
8,9
the level of NET), arousal, anxiety, attention-defi-
cit/hyperactivity disorder (atomoxetine 1 is the first non-
stimulant selective NET inhibitor for the treatment of
1
1
17
[ C]Nisoxetine 2 showed significant binding in NET-
rich tissue in mouse brain but also a high nonspecific
binding and unfavorable slow kinetics, which excludes
it as a practical NET imaging agent. Recently, several
research groups have reported the evaluation of deriva-
tives of reboxetine (3) with more promising properties
1
0
11
12
ADHD ), and Alzheimer’s disease. Therefore, the
Keywords: SPECT; Imaging; Norepinephrine; Transporter; Reboxe-
tine; Iodine-123.
q
Preliminary results of these experiments have been reported at
scientific conferences for imaging 1. See Refs. 1–3.
1
1
18–20
than the nisoxetine series. [ C]MeNER
4 has
shown the highest hypothalamus/striatum ratio of 2.5
at 60 min in rats, corresponding to the relative
*
Corresponding author. Tel.: +1 203 401 4309; fax: +1 203 789
119; e-mail: gtamagnan@indd.org
2
1
2
distribution of NET and SERT. Displacement study
0
960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.018

534
G. D. Tamagnan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 533–537
using reboxetine or desipramine as selective NET inhib-
itor pretreatment reduced this ratio to unity, whereas
selective DAT and SERT inhibitors did not significantly
SERT, which, when combined with the significantly
higher density of SERT compared to NET, decreased
the selectivity for NET in vivo.
1
8
affect the ratio. However, the long time to achieve
peak uptake (>90 min) makes this radioligand unsuit-
2
We recently reported a new asymmetric synthesis which
1
9
able for PET imaging. From a statistical point of view
it is not optimal to obtain the maximal specific binding
at the end of the experiment when the signal to noise ra-
tios are decreasing due to rapid decay of the carbon-11
we used to synthesize new analogs of reboxetine. We re-
port here the synthesis, in vitro evaluation of new
reboxetine analogs as potential NET imaging agent as
well as the in vivo evaluation of the most promising
agent in this series (Fig. 1).
(
ical for SPECT imaging owing to the longer half-life of
half-life = 20 min), although these issues are not as crit-
123
typical SPECT radionuclides ( I, 13.2 h,
111
In, 2.8
days). In the nisoxetine series, (R)-[ I]MIPP 5 showed
good in vitro affinity for NET and moderate affinity for
Since in the reboxetine series, the most active isomer al-
ways corresponds to the S,S isomer, in order to rapidly
discriminate the future candidates we decided to first
synthesize the racemic compounds (mixture S,S/R,R)
and then to pursue the most promising candidates with
the evaluation of the S,S enantiomer. The synthesis of
the new compounds is summarized in Scheme 1. The
key intermediate 11 was synthesized in good yield start-
1
25
R
O
R
O
O
S
S
ing from the racemic aminopropanediol 6 in five steps
2
N
R
H
(detail of the synthesis was described previously ). Sub-
N
H
sequent condensation of substituted diaryl or diheteroa-
ryl zinc provided a mixture of racemic diastereomers 12
and 13 easily separated by chromatography. The race-
mic (S,S/R,R) reboxetine analogs of N-Boc-protected
14a–i were obtained from 12 and 13, using either a Mits-
unobu reaction (starting from 12) or by a direct nucleo-
1
2
5
R = Me Atomoxetine
R = OMe Nisoxetine
3
4
R = OEt Reboxetine
R = Me MeNER
R = I
MIPP
Figure 1. Structure of NET ligand.
HO
a
OH
O
HO
OH
O
b
HO
c
f
NH Cl
O
NH2
N
O
H
6
7
8
O
OHC
O
HO
HO
d
e
N
N
N
H
9
Boc
10
Boc
11
OH
OH
O
N
O
Ar
+
Ar
I
S
I
Ar =
N
Boc
Rac 2S,3R/2R,3S
Boc
Rac 2S,3S/2R,3R
1
2
13
Substitent
Cr(CO)
g
h
CH
O
3
CF
3
F
3
O
R
R =
R
O
I
O
O
i
O
Ar
Ar
O
NO
2
N
N
H
Boc
Rac 2S,3S/2R,3R
Rac 2S,3S/2R,3R
1
4a-i
15a-i
Scheme 1. Reagents and conditions: (a) ClCH COCl/Et
2
3
N; (b) t-BuOK/t-AmOH; (c) red-Al/THF; (d) (Boc)
2
O/NaOH; (e) trichloroisocyanuric acid/
Cl
TEMPO/NaHCO ; (f) Ar Zn/THF; (g) R-OH/DIAD/PPh
3
2
3
/THF; (h) fluorobenzenetricarbonylchromium complex/NaH; (i) TFA/CH
2
2
.

G. D. Tamagnan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 533–537
535
a
i
Table 1. Affinity (K , nM) or SERT inhibition (% at 10 lM) of reboxetine analogs
b
Compound
Structure
NET
DAT
SERT
SERT inhibition (%)
I
O
O
1
1
1
1
1
1
1
1
1
5a
5b
5c
5d
5e
5f
2.47
410
40
12
24
44
N
H
O
O
O
I
I
I
I
15.4
27.6
28.5
101
>10,000
N
H
CH3
O
O
740
N
H
O
O
O
3139
1871
14.2
716
N
H
CF3
O
O
486
252
351
N
H
NO2
O
O
O
70.4
73.8
71.9
79.6
I
I
N
H
O
O
5g
5h
5i
N
H
CH3
O
186
31
S
S
O
I
N
H
O
O
210
141
O
I
N
H
a
Assays of novel compounds employed cell membrane-containing homogenates of rat forebrain (frontoparietal cerebral cortex for SERT and NET,
3
caudate-putamen for DAT), with [ H]paroxetine (0.17 nM) for SERT (blank = 10mM R,S-fluoxetine, Sigma-RBI), [ H]GBR-12935 (0.20 nM) for
3
3
DAT (blank = 10 mM GBR-12909, Sigma-RBI), and [ H]nisoxetine (0.40 nM) for NET (blank = 10 mM desipramine, Sigma-RBI); results are
expressed as means of three separate determinations.
All compounds are racemic mixtures; for clarity, just one enantiomer is represented.
b

536
G. D. Tamagnan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 533–537
a
Table 2. Binding affinity of INER (K
i
, nM, means ± SEM)
Compound
INER
NET
DAT
SERT
Selectivity NET versus DAT
270
Selectivity NET versus SERT
51
0.84 ± 0.12
228 ± 6.9
40.4 ± 10.4
a
Assays of INER employed cell membrane-containing homogenates of rat forebrain (frontoparietal cerebral cortex for SERT and NET, caudate-
3
3
putamen for DAT), with [ H]paroxetine (0.17 nM) for SERT (blank = 10 mM R,S-fluoxetine, Sigma-RBI), [ H]GBR-12935 (0.20 lM) for DAT
3
(
blank = 10 lM GBR-12909, Sigma-RBI), and [ H]nisoxetine (0.40 nM) for NET (blank = 10 lM desipramine, Sigma-RBI); results are expressed
as means of three separate determinations.
philic substitution of the appropriate 2-fluoro benzene
tricarbonylchromium complex (starting from 13).
Deprotection of compounds 14a–i was achieved in
quantitative yield using trifluoroacetic acid. Compounds
These results indicate that compound 15a shows the
most promise and thus we focused our effort on this
structure. The enantiomeric pure synthesis was achieved
using the same sequence illustrated in Scheme 1 but this
time starting from the chiral (S)-(-)-3-amino-1,2-pro-
panediol which led to INER (formal (2S,3S)-15a) with
ee = 99%. INER was evaluated against all three mono-
amine transporters and shows binding value
1
5a–i were evaluated in vitro against all three mono-
amine transporters, using cell membrane-containing
homogenates of rat forebrain (frontoparietal cerebral
cortex for SERT and NET, caudate-putamen for
DAT) as transporter sources (Table 1).
K = 0.84 nM for NET and selectivity versus DAT and
i
SERT (270 and 51, respectively, Table 2).
All compounds showed a better activity at NET than at
the other monoamine transporters (DAT and SERT) ex-
cept for compound 15f, which was slightly more active
at the DAT. In this series, compound 15a is of particular
interest, with a binding affinity of 2.5 nM for the racemic
mixture. The introduction of an iodine at position 3 of
the phenyl moiety (compounds 15b–e) led to a decrease
of the affinity. Replacement of the phenoxy by a benzyl-
oxy moiety (15f–g) led to less active compounds
1
23
Radiolabeled [ I]INER was prepared via iododestan-
nylation of the tin precursor 16, easily synthesized by
palladium-catalyzed stannylation of Boc-INER (formal
(2S,3S)-14a). Removal of the Boc-protecting group with
trifluoroacetic acid and HPLC purification led to
1
23
[
I]INER as described in Scheme 2. The lipophilicity
of [ I]INER was assessed by octanol-phosphate buffer
(pH 7.4) partition. The partition coefficient logD of
[ I]INER was 2.6, which is high in the range of values
considered acceptable for good blood–brain barrier
1
23
1
23
(
15.4 nM vs 73.8 nM for 15b vs 15g). Replacement of
the phenyl ring by a thiophene reduced the activity of
the compounds compared to the parent.
2
2
penetration.
123I
I
SnMe3
O
O
O
a
b
O
O
O
N
N
N
Boc
Boc
Boc
[123I]INER
Boc-INER
16
123
/DME; (b) Na[ I]/CH
Scheme 2. Reagents and conditions: (a) (SnMe
3
)
2
/Pd(PPh
3
)
4
3
COOOH/H
3
4
PO .
1
1
2.5
0.0
Diencephalon
Temporal Cortex
Parietal Cortex
Brain Stem
Frontal Cortex
Occipital Cortex
Cerebellum
7
.5
.0
.5
.0
5
2
0
0
1
2
3
4
5
6
TIME (h)
1
23
Figure 2. Decay-corrected time activity curve in a baboon undergoing constant infusion of [ I]INER.

G. D. Tamagnan et al. / Bioorg. Med. Chem. Lett. 17 (2007) 533–537
537
A series of SPECT imaging studies were conducted
including both bolus-only and bolus plus continuous
3. Brenner, E.; Baldwin, R. M.; Tarazi, F.; Baldessarini, R.
J.; Tamagnan, G. D. The 229th ACS National Meeting,
San Diego, CA, March 13–17, 2005.
1
23
infusion of [ I]INER in an ovariectomized female ba-
boon (Papio anubis) to evaluate the regional brain up-
take and washout of activity. Pharmacological
specificity was assessed by blocking displacement exper-
iments using reboxetine as selective NET antagonist and
citalopram as SERT antagonist.
4
. Aston-Jones, G.; Foote, S.; Bloom, F. Anatomy and
physiology of locus coeruleus neurons: functional impli-
cation. In Norepinephrine (Frontiers of Clinical Neurosci-
ence); Ziegler, M., Lake, C., Eds.; Williams and Wilkins:
Baltimore, 1984; Vol. 2, pp 92–116.
5
. Spillmann, M. K.; Van der Does, A. J. W.; Rankin, M. A.;
Vuolo, R. D.; Alpert, J. E.; Nierenberg, A. A.; Rosen-
baum, J. F.; Hayden, D.; Schoenfeld, D.; Fava, M.
Psychopharmacology 2001, 155, 123.
An initial study was performed by dynamic SPECT
imaging after a single bolus injection of 7.89 mCi of
1
23
6
7
. Miller, H. L.; Delgado, P. L.; Salomon, R. M.; Berman,
R.; Krystal, J. H.; Heninger, G. R.; Charney, D. S. Arch.
Gen. Psychiatry 1996, 53, 117.
. Kugaya, A.; Sanacora, G.; Staley, J. K.; Malison, R. T.;
Bozkurt, A.; Khan, S.; Anand, A.; van Dyck, C. H.;
Baldwin, R. M.; Seibyl, J. P. Biol. Psychiatry 2004, 56,
[
I]INER in an adult female olive baboon. The maxi-
mal radioactivity peak in the brain was obtained after
0 min, with 1% of the total injected dose, followed by
1
a slow washout over more than 2 h. Regional brain trac-
er uptake is consistent with the distribution of NET in
the baboon brain, with the highest uptake in locus coe-
ruleus (brainstem) and thalamus (diencephalon), and
lowest uptake in cerebellum and striatum. In the next
4
97.
8
. Hisamichi, S.; Moon, A. Y.; Aoyagi, A.; Hara, T.; Abe,
K.; Yamazaki, R.; Kumagae, Y.; Naruto, S.; Koyama, K.;
Marumoto, S.; Tago, K.; Toda, N.; Takami, K.; Yamada,
N.; Ori, M.; Kogen, H.; Kaneko, T. J. Pharmacol. Sci.
1
23
study [ I]INER was administered as a loading bolus
injection followed by constant infusion using a bolus
to infusion ratio of 1.3 h (Fig. 2), stable levels were ob-
tained in less than 100 min. Regional distribution of
radioactivity was identical to that obtained in the bolus
experiment, which followed the established distribution
of NET in the baboon brain.
2
003, 93, 95.
. Allgulander, C.; Hackett, D.; Salinas, E. Br. J. Psychiatry
001, 179, 15.
9
2
1
0. Bymaster, F. P.; Katner, J. S.; Nelson, D. L.; Hemrick-
Luecke, S. K.; Threlkeld, P. G.; Heiligenstein, J. H.;
Morin, S. M.; Gehlert, D. R.; Perry, K. W. Neuropsycho-
pharmacology 2002, 27, 699.
To assess the pharmacological specificity, reboxetine was
injected (1 mg/kg iv) at 210 min after administration of
bolus-plus-constant infusion which resulted in a reduc-
tion of radioactivity in all brain regions. Reduction was
more pronounced in the NET-rich regions, with up to
1
1. Biederman, J.; Spencer, T. Biol. Psychiatry 1999, 46,
1234.
12. Tejani-Butt, S. M.; Yang, J.; Zaffar, H. Brain Res. 1993,
631, 147.
1
3. Scanley, B. E.; Al-Tikriti, M. S.; Gandelman, M. S.;
Laruelle, M.; Zea-Ponce, Y.; Baldwin, R. M.; Zoghbi, S.
S.; Hoffer, P. B.; Charney, D. S.; Wang, S.-Y.; Neumeyer,
J. L.; Innis, R. B. Eur. J. Nucl. Med. 1995, 22, 4.
6
citalopram (5 mg/kg) or methylphenidate (0.5 mg/kg) at
0% specific displacement. In contrast, displacement with
123
10 min pi did not show any displacement of [ I]INER.
2
1
4. Malison, R. T.; McDougle, C. J.; van Dyck, C. H.; Scahill,
L.; Baldwin, R. M.; Seibyl, J. P.; Price, L. H.; Leckman, J.
F.; Innis, R. B. Am. J. Psychiatry 1995, 152, 1359.
Recently, the same structure as compound 15a was
2
3
reported under the name ‘IPBM’ by Saji et al. Results
were comparable to those we have reported here.
1
5. Chen, P.; Kilts, C. D.; Galt, J. R.; Ely, T. D.; Camp, V.
M.; Malveaux, E. J.; Goodman, M. M. J. Nucl. Med.
2
000, 41, 39P.
In conclusion, we report the synthesis of nine racemic po-
tential NET imaging agents. Of these, compound 15a
16. Goodman, M. M.; Chen, P.; Plisson, C.; Martarello, L.;
Galt, J.; Votaw, J. R.; Kilts, C. D.; Malveaux, G.; Camp,
V. M.; Shi, B.; Ely, T. D.; Howell, L.; McConathy, J.;
Nemeroff, C. B. J. Med. Chem. 2003, 46, 925.
(
INER) was the best candidate for further investigation
and we developed an asymmetric synthesis leading to
the enantiopure INER. INER showed high affinity for
NET (K = 0.84 nM) and good selectivity versus DAT
1
7. Haka, M. S.; Kilbourn, M. R. Nucl. Med. Biol. 1989, 16,
71.
7
i
1
8. Wilson, A. A.; Patrick Johnson, D.; Mozley, D.; Hussey,
D.; Ginovart, N.; Nobrega, J.; Garcia, A.; Meyer, J.;
Houle, S. Nucl. Med. Biol. 2003, 30, 85.
and SERT (270 and 51, respectively). We synthesized
1
23
[
baboon. Taken together, the in vivo specificity, selectivi-
I]INER and showed specific in vivo NET binding in
19. Schou, M.; Halldin, C.; Sovago, J.; Pike, V. W.; Gulyas,
B.; Mozley, P. D.; Johnson, D. P.; Hall, H.; Innis, R. B.;
Farde, L. Nucl. Med. Biol. 2003, 30, 707.
1
23
ty, and kinetics suggest that [ I]INER is a promising
agent for imaging NET in vivo using SPECT or PET.
2
0. Schou, M.; Halldin, C.; S o´ v a´ g o´ , J.; Pike, V. W.; Hall, H.;
Guly a´ s, B.; Mozley, P. D.; Dobson, D.; Shchukin, E.;
Innis, R. B.; Farde, L. Synapse 2004, 53, 57.
1. Smith, H. R.; Beveridge, T. J. R.; Porrino, L. J. Neuro-
science 2006, 138, 703.
2. Laruelle, M.; Slifstein, M.; Huang, Y. Mol. Imaging Biol.
References and notes
2
2
2
1
. Tamagnan, G. D.; Haile, C. N.; Brenner, E.; Koren, A.
O.; Bois, F.; Amici, L.; Baldwin, R. M.; Seibyl, J. P.;
Staley, J. K. EJNM 2005, 32, 287.
. Brenner, E.; Baldwin, R. M.; Tamagnan, G. D. Org. Lett.
2
2
003, 5, 363.
3. Kanegawa, N.; Kiyono, Y.; Kimura, H.; Sugita, T.;
Kajiyama, S.; Kawashima, H.; Ueda, M.; Kuge, Y.; Saji,
H. EJNM 2006, 33, 639.
2
005, 7, 937.