Novel quinoxaline derivatives for in vivo imaging of β-amyloid plaques in the brain

Article abstract of DOI:10.1016/j.bmcl.2011.05.079

In a search for new probes to detect β-amyloid plaques in the brain of patients with Alzheimer's disease (AD), we have synthesized and evaluated a series of quinoxaline derivatives containing a '6+6-6' ring system. These quinoxaline derivatives showed excellent affinity for Aβ_1-42 aggregates with K_i values ranging from 2.6 to 10.7 nM. Autoradiography with sections of brain tissue from an animal model of AD mice (APP/PS1) and AD patients revealed that [¹²⁵I]5 labeled β-amyloid plaques specifically. In biodistribution experiments using normal mice, [¹²⁵I]5 displayed high uptake (6.03% ID/g at 2 min) into and a moderately fast washout from the brain. Although additional refinements are needed to decrease the lipophilicity and improve the washout rate, the quinoxaline scaffold may be useful as a backbone structure to develop novel β-amyloid imaging agents.

Full text of DOI:10.1016/j.bmcl.2011.05.079

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4193–4196
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel quinoxaline derivatives for in vivo imaging of b-amyloid plaques
in the brain
Mengchao Cui a,b, Masahiro Ono a, , Hiroyuki Kimura a, Boli Liu b, Hideo Saji a,
⇑
⇑
a Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida Shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan
b Key Laboratory of Radiopharmaceuticals, Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
In a search for new probes to detect b-amyloid plaques in the brain of patients with Alzheimer’s disease
(AD), we have synthesized and evaluated a series of quinoxaline derivatives containing a ‘6+6ꢀ6’ ring
system. These quinoxaline derivatives showed excellent affinity for Ab1–42 aggregates with Ki values
ranging from 2.6 to 10.7 nM. Autoradiography with sections of brain tissue from an animal model of
AD mice (APP/PS1) and AD patients revealed that [125I]5 labeled b-amyloid plaques specifically. In biodis-
tribution experiments using normal mice, [125I]5 displayed high uptake (6.03% ID/g at 2 min) into and a
moderately fast washout from the brain. Although additional refinements are needed to decrease the
lipophilicity and improve the washout rate, the quinoxaline scaffold may be useful as a backbone struc-
ture to develop novel b-amyloid imaging agents.
Received 17 March 2011
Revised 20 May 2011
Accepted 23 May 2011
Available online 27 May 2011
Keywords:
Alzheimer’s disease
b-Amyloid
Imaging agent
Binding assay
Autoradiography
Ó 2011 Elsevier Ltd. All rights reserved.
Alzheimer’s disease (AD) is the most common form of dementia
in older people, and is estimated to account for 50% to 70% of
dementia cases. Senile plaques containing b-amyloid (Ab) aggre-
gates and neurofibrillary tangles composed of highly phosphory-
lated tau protein are the two key pathological findings in
postmortem AD brains.1,2 The precise molecular mechanisms lead-
ing to the development of this disease are not fully understood,
however, the amyloid cascade hypothesis is the most accepted
explanation for the pathogenesis of AD.3
With modern functional neuroimaging techniques such as pos-
itron emission tomography (PET) and single photon emission
tomography (SPECT), radio-labeled probes that specifically target
Ab plaques may aid in the diagnosis and monitoring of AD in living
subjects. Thus, many efforts have focused on developing radiotra-
cers or agents that allow Ab imaging in vivo. An uncharged Thiofla-
vin T (ThT) analogue, [11C]PIB (2-(40-[11C]methylaminophenyl)-6-
hydroxybenzothiazole), which is highly efficient both in crossing
the blood–brain barrier (BBB) and in selective binding to Ab pla-
ques, is the most widely used amyloid marker with PET for amyloid
imaging.4,5 Recently, a close analogue of PIB, [11C]AZD2184 ([11C]-
Compared with 11C (t1/2: 20 min), 18F with a longer half-life (t1/2
:
110 min) would be more useful for routine clinical use. In fact, a
18F-labeled PIB analogue, [18F]GE-067 ([18F]-2-(3-fluoro-4-(meth-
ylamino)phenyl)benzo[d]thiazol-6-ol)8, and two stilbene deriva-
tives with fluorinated polyethylene glycol (PEG) units,
[
18F]BAY94-9172 ([18F]-4-(N-methylamino)-40-(2-(2-(2-fluoroeth-
oxy)ethoxy)ethoxy)-stilbene)9 and [18F]AV-45 ([18F]-(E)-4-(2-(6-
(2-(2-(2-fluoroethoxy)ethoxy)ethoxy)pyridin-3-yl)vinyl)-N-methy-
laniline)10, are currently under phase II or phase III clinical trials.
However, since the failure of [
123I]IMPY ([123I]-6-iodo-2-(40-
dimethylamino-)phenyl-imidazo[1,2]pyridine) in clinical trials,
there has been no report of any Ab imaging agents for SPECT mov-
ing into clinical testing.11–13
Most of these Ab probes were derived from the basic structure
of thioflavin-T, which consists of a six-membered ring fused to a
X
R1
6
R2
6
5
6
R2
6
6
Y
R1
"6+6-6" ring system
"6+5-6" ring system
2-[6-(methylamino)pyridin-3-yl]-1,3-benzothiazol-6-ol), with
a
high affinity for Ab fibrils in vitro was reported. Preliminary PET
studies indicated that this probe displayed fast kinetics and an
excellent contrast for Ab plaques in the brain of AD patients.6,7
HN
HN
F
R1
N
O
3
R2
Napthphen
N
Quinoxaline
F
3
⇑
Corresponding authors. Tel.: +81 75 753 4608; fax: +81 75 753 4568 (M.O.); tel.:
O
+81 75 753 4556; fax: +81 75 753 4568 (H.S.).
E-mail addresses: ono@pharm.kyoto-u.ac.jp (M. Ono), hsaji@pharm.kyoto-u.ac.jp
(H. Saji).
Phenapth
Figure 1. Design strategy for the quinoxaline derivatives as probes for Ab plaques.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.05.079

4194
M. Cui et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4193–4196
Scheme 1. Reagents and conditions: (a) 1: conc. H2SO4, Br2, 0 °C to rt 2: ethyl phosphite, Et3N, THF, 0 °C to rt; (b) 4-bromobenzene-1,2-diamine, DMSO, rt; (c) (Bu3Sn)2,
(PPh3)4Pd, dioxiane, Et3N, reflux; (d) [125I]NaI, HCl (1 M), H2O2 (3%), rt; (e) I2, CHCl3, rt; (f) Br2, Et2O, 0 °C to rt; (g) SnCl2, EtOH, conc. HCl, reflux; (h) DMSO, rt; (i) acetone, CH3I,
K2CO3.
five-membered heterocyclic ring together with another conjugated
six-membered ring, creating a ‘6+5ꢀ6’ ring system (Fig. 1). In an at-
120
3
5
11
12
100
80
60
40
20
0
tempt to further develop novel ligands for the imaging of amyloid
plaques through enlargement of the five-membered ring, Zhang
et al. reported two fluoro-pegylated phenapth or napthphen deriv-
atives as PET probes which contain a ‘6+6ꢀ6’ ring system.14 How-
ever, the naphthalene ring is highly lipophilic which may cause
high non-specific binding. To reduce the lipophilicity, we have se-
lected quinoxaline instead of the naphthalene ring. Herein, we re-
port the synthesis and biological evaluation of quinoxaline
derivatives as potential Ab imaging agents.
The synthesis of quinoxaline derivatives is described in
-12 -11 -10 -9
-8
-7
-6
-5
-4
Scheme 1, starting from
a-bromoacetophenone and substituted
Log [Inhibitor]
Figure 3. Inhibition curves for the binding of [125I]IMPY to Ab1–42 aggregates.
o-phenylenediamines. The intermediates
a-bromoacetophenone
(2 and 7) and 4-iodobenzene-1,2-diamine (9) were prepared first
based on procedures reported previously.15 In the presence of
DMSO, the quinoxaline backbone was formed by an one-pot tan-
dem oxide condensation procedure in moderate yields (3: 36.7%,
10: 49.1%). Two isomeric products were obtained depending on
the course of cyclization, the corresponding 2-aryl-6-substitued
quinoxalines (3 and 10) as major products and the 2-aryl-7-sub-
stitued quinoxalines as minor products. Furthermore, the crystal
structure of 3 was determined by X-ray crystallography (see Sup-
plementary data). The free amino derivative 11 was achieved by
reducing a nitro group to an amino group with SnCl2 in ethanol un-
der reflux condition (87.0%). Conversion of 11 to the N-methyl-
amino derivative 12 was achieved by methylation with CH3I
under alkaline conditions. The desired tributyltin precursor 4 was
prepared by Pd(PPh3)4-catalyzed trans-stannylation from the bro-
mide compound 3 (31.5%). The subsequent iododestannylation
reaction afforded the iodinated target 5 (21.8%).
Figure 2. HPLC profiles of 5 (A) and [125I]5 (B). HPLC conditions: Cosmosil C18
column (Nakalai Tesque, 5C18-AR-II, 4.6 mm ꢁ 150 mm), CH3CN/H2O = 90/10, 1 mL/
min, UV, 254 nm, 5tR (UV) = 8.54 min, [125I]5tR (RI) = 8.65 min.
The desired radioiodinated ligand [125I]5 was successfully pre-
pared from the corresponding tributyltin precursors through

M. Cui et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4193–4196
4195
Table 1
aggregates (Ki 610 nM). The results indicated that these quinoxa-
line derivatives bind to the same site as [125I]IMPY. From the Ki val-
ues listed in Table 1, it is clear that the tertiary N,N-dimethylamino
analogues 3 and 5 have higher affinity (Ki = 2.6 and 4.1 nM, respec-
tively) than the secondary N-methylamino analogue 12
(Ki = 7.7 nM), while the corresponding primary amino analogue
11 showed moderate affinity (Ki = 10.7 nM), which is consistent
with previous data on primary, secondary, and tertiary amino li-
gands.16 The Ki values of these quinoxaline derivatives are similar
to that of phenapth or napthphen derivatives (Ki = 1.6 nM and
3.0 nM) reported previously.14 Due to the encouraging binding
data for these quinoxaline compounds, the tertiary N,N-dimethyl-
amino analogue (5) was selected for radiolabeling and further bio-
logical evaluation.
To confirm the specific binding of 5 to Ab plaques, the fluores-
cent staining of sections of brain tissue from an animal model of
AD (C57BL6, APP/PS1 mouse, 12 months) was performed. As
shown in Figs. 4, 5 clearly labeled Ab plaques with low background
levels (Fig. 4A), the labeling pattern consistent with that obtained
on staining with thioflavin-S in adjacent sections (Fig. 4B).
Next, in vitro autoradiography was carried out in brain sections
of AD patients and double transgenic mice using the radioiodinated
Inhibition constants (Ki, nM) for the binding of [125I]IMPY to Ab1–42 aggregates.
Compound
R1
R2
Ki (nM)a
3
5
Br
I
I
I
—
N(CH3)2
N(CH3)2
NH2
NHCH3
—
2.6 0.2
4.1 0.7
10.7 1.1
7.7 1.4
10.5 1.0
11
12
IMPY
a
Values are the mean for three independent experiments.
standard iododestannylaltion reactions, using sodium [125I]iodide,
hydrogen peroxide, and hydrochloric acid (Scheme 1). The overall
radiochemical yield for [125I]5 was 62.3%. The radiochemical iden-
tity of the 125I-labeled ligands was confirmed by co-injection with
nonradioactive compounds from HPLC profiles (Fig. 2). The purified
ligands all had a radiochemical puritiy greater than 98% and high
specific activity (no carrier added, approx. 81.4 TBq/mmol).
The affinity (Ki, nM) of the quinoxaline compounds was evalu-
ated based on inhibition of the binding of [125I]IMPY to aggregates
of Ab1–42 fibers in solution (Fig. 3). The results are presented in
Table 1. All of the compounds showed excellent affinity for Ab
Figure 4. Fluorescence staining of 5 (A) in a section of brain tissue from a Tg model mouse (C57BL6, APP/PS1, 12 months old, male) with a filter set for GFP. The presence of
plaques was confirmed by staining of the adjacent section with thioflavin S (B).
Figure 5. Autoradiography of [125I]5 in vitro in brain sections of an AD patient (A) and Tg mouse (C57BL6, APP/PS1, 12 months old, male) (C). The presence and distribution of
plaques in the sections were confirmed with immunohistochemical staining using a monoclonal Ab antibody (B) and thioflavin-S (D). Arrows show the correspondence to Ab
plaques.

4196
M. Cui et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4193–4196
Table 2
was washed out from the normal mouse brain moderately quickly.
Further refinements are under way to decrease the lipophilicity of
this probe and improve its rate of washout from the brain. These
quinoxaline ligands could be potentially useful for the imaging of
Ab plaques in living subjects.
Biodistribution in normal ddy mice after iv injection of [125I]5a and the lipophilicity
(log D) of the ligand
[
125I]5 Log D = 4.02 0.12
Organ
2 min
30 min
60 min
120 min
Blood
Brain
Heart
Liver
Spleen
Lung
Kidney
Stomachb
Intestine
4.71 1.04
6.03 0.99
14.91 2.11
30.49 7.74
5.69 1.79
14.53 5.53
19.67 4.99
2.06 0.39
3.43 0.83
2.88 0.62
5.21 1.16
2.53 0.77
15.73 3.76
2.82 0.76
3.99 1.08
8.76 2.72
3.28 0.72
18.62 12.17
2.98 1.04
2.91 1.18
2.17 0.69
12.50 5.00
2.22 0.86
3.53 1.27
6.68 2.53
5.39 1.66
22.50 9.17
1.88 0.55
1.12 0.31
1.22 0.29
7.53 2.51
1.17 0.23
1.95 0.42
3.96 0.96
8.62 2.40
16.93 5.92
Acknowledgements
This study was supported by the Japan Society for the Promo-
tion of Science (JSPS) through the ‘‘Funding Program for Next Gen-
eration World-Leading Researchers (NEXT Program)’’ initiated by
the Council for Science and Technology Policy (CSTP), and a
Grant-in-aid for Young Scientists (A) and Exploratory Research
from the Ministry of Education, Culture, Sports, Science and Tech-
nology, Japan. This study was also supported by the China Scholar-
ship Council (CSC).
a
Expressed as
deviation.
% injected dose per gram. Average for five mice standard
b
Expressed as % injected dose per organ.
Supplementary data
ligand [125I]5. As shown in Figure 5A, specific labeling of plaques
was observed in AD brains. Immunohistochemical staining con-
firmed the presence of plaques in these sections (Fig. 5B). As ex-
pected, [125I]5 also labeled Ab plaques in the brain sections of
transgenic mice, with numerous fluorescent spots in the cortex re-
gion, and minimal background noise (Fig. 5 C). The distribution of
Ab plaques was consistent with the results of autoradiography
with thioflavin-S (Fig. 5D).
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.05.079.
References and notes
1. Hardy, J. A.; Higgins, G. A. Science 1992, 256, 184.
2. Selkoe, D. J. JAMA-J. Am. Med. Assoc. 2000, 283, 1615.
3. Hardy, J. A.; Selkoe, D. J. Science 2002, 297, 353.
Although two nitrogen atoms were introduced into the aro-
matic system of the quinoxaline structure, a high log D value
(4.02) for [125I]5 was obtained under the experimental conditions,
a reflection of the lipophilic properties of the radioiodinated ligand.
Nonetheless, biodistribution experiments in normal mice (Table 2)
clearly indicated that [125I]5 readily penetrated the intact BBB
showing excellent initial uptakes into the brain (6.03 0.99% ID/
g) at 2 min after injection. Since there are no Ab plaques in normal
mice, the high uptake was subsequently followed by a moderate
washout with 1.12% ID/g at 120 min. Compared with the fluoro-
pegylated phenapth derivatives (3.80)14, the radioiodinated ligand
4. Mathis, C. A.; Wang, Y.; Holt, D. P.; Huang, G. F.; Debnath, M. L.; Klunk, W. E. J.
Med. Chem. 2003, 46, 2740.
5. Klunk, W. E.; Engler, H.; Nordberg, A.; Wang, Y.; Blomqvist, G.; Holt, D. P.;
Bergstrom, M.; Savitcheva, I.; Huang, G. F.; Estrada, S.; Ausen, B.; Debnath, M.
L.; Barletta, J.; Price, J. C.; Sandell, J.; Lopresti, B. J.; Wall, A.; Koivisto, P.; Antoni,
G.; Mathis, C. A.; Langstrom, B. Ann. Neurol. 2004, 55, 306.
6. Andersson, J. D.; Varnas, K.; Cselenyi, Z.; Gulyas, B.; Wensbo, D.; Finnema, S. J.;
Swahn, B. M.; Svensson, S.; Nyberg, S.; Farde, L.; Halldin, C. Synapse 2010, 64,
733.
7. Nyberg, S.; Jonhagen, M. E.; Cselenyi, Z.; Halldin, C.; Julin, P.; Olsson, H.;
Freund-Levi, Y.; Andersson, J.; Varnas, K.; Svensson, S.; Farde, L. Eur. J. Nucl. Med.
Mol. Imaging 2009, 36, 1859.
8. Koole, M.; Lewis, D. M.; Buckley, C.; Nelissen, N.; Vandenbulcke, M.; Brooks, D.
J.; Vandenberghe, R.; Van Laere, K. J. Nucl. Med. 2009, 50, 818.
9. Rowe, C. C.; Ackerman, U.; Browne, W.; Mulligan, R.; Pike, K. L.; O’Keefe, G.;
Tochon-Danguy, H.; Chan, G.; Berlangieri, S. U.; Jones, G.; Dickinson-Rowe, K.
L.; Kung, H. F.; Zhang, W.; Kung, M. P.; Skovronsky, D.; Dyrks, T.; Holl, G.;
Krause, S.; Friebe, M.; Lehman, L.; Lindemann, S.; Dinkelborg, L. M.; Masters, C.
L.; Villemagne, V. L. Lancet Neurol. 2008, 7, 129.
10. Choi, S. R.; Golding, G.; Zhuang, Z. P.; Zhang, W.; Lim, N.; Hefti, F.; Benedum, T.
E.; Kilbourn, M. R.; Skovronsky, D.; Kung, H. F. J. Nucl. Med. 2009, 50, 1887.
11. Kung, M. P.; Hou, C.; Zhuang, Z. P.; Zhang, B.; Skovronsky, D.; Trojanowski, J. Q.;
Lee, V. M.; Kung, H. F. Brain Res. 2002, 956, 202.
[
125I]5 in this study had superior brain2 min/brain120 min ratios
(5.38). In addition, the blood background level during the experi-
ment was lower than the brain uptake, which is good for reducing
nonspecific binding. [125I]5 was excreted predominantly by the
hepatobiliary and excretory systems, and radioactivity was ob-
served to accumulate within the intestine at later time points
(16.93% ID/g at 120 min).
12. Zhuang, Z. P.; Kung, M. P.; Wilson, A.; Lee, C. W.; Plossl, K.; Hou, C.; Holtzman,
D. M.; Kung, H. F. J. Med. Chem. 2003, 46, 237.
13. Newberg, A. B.; Wintering, N. A.; Plossl, K.; Hochold, J.; Stabin, M. G.; Watson,
M.; Skovronsky, D.; Clark, C. M.; Kung, M. P.; Kung, H. F. J. Nucl. Med. 2006, 47,
748.
A good initial uptake combined with a rapid washout from nor-
mal brain tissue are important requirements for Ab plaque imaging
agents.17 The radioiodinated quinoxaline derivative [125I]5 re-
ported here, met the first requirement but still needs some refine-
ments to improve its washout rate.
14. Zhang, W.; Kung, M.-P.; Hou, C.; Oya, S.; Skovronsky, D.; Manchanda, R.; Choi,
S.-R.; Kung, H. J. Nucl. Med. 2008, 49, 143P.
In summary, we have developed a new type of amyloid imaging
agent based on the quinoxaline pharmacophore. These derivatives
displayed high affinity for Ab1–42 aggregates (Ki values in the nM
range). The radioiodinated probe [125I]5 showed specific labeling
of Ab plaques in sections of brain tissue from AD patients and
transgenic mice. In addition, [125I]5 readily enters the brain and
15. Hindo, S. S.; Mancino, A. M.; Braymer, J. J.; Liu, Y.; Vivekanandan, S.;
Ramamoorthy, A.; Lim, M. H. J. Am. Chem. Soc. 2009, 131, 16663.
16. Cai, L. S.; Innis, R. B.; Pike, V. W. Curr. Med. Chem. 2007, 14, 19.
17. Kung, H. F.; Choi, S. R.; Qu, W. C.; Zhang, W.; Skovronsky, D. J. Med. Chem. 2010,
53, 933.