Synthesis and preliminary evaluation of 2-arylhydroxyquinoline derivatives for tau imaging

Article abstract of DOI:10.1002/jlcr.3133

Alzheimer's disease (AD) is the most common cause of dementia. Senile plaques, consisting of β-amyloid, and neurofibrillary tangles (NFTs), composed of tau protein, are representative pathological hallmarks of AD. It is believed that the accumulation of NFTs precedes the onset of clinical symptoms of AD and correlates with the progression of memory dysfunction. Thus, the use of noninvasive detection techniques including radiolabeled probes and positron emission tomography (PET) will facilitate early diagnosis or staging of AD. In this study, we synthesized and evaluated novel hydroxylated 2-arylquinoline derivatives as tau imaging PET probes. The binding affinities of compounds for tau were evaluated by fluorescent staining of the AD hippocampal section and a competitive binding assay using [¹⁸F]THK-523. THK-951 showed high binding affinity for tau pathology in an AD brain section and K18Δ280K fibrils (K_i = 20.7 nM); thus, we radiosynthesized a ¹¹C-labeled THK-951 and further studied its potential as a tau PET probe. The [¹¹C]THK-951 demonstrated excellent kinetics in a normal mouse brain (3.23% ID/g at 2 min postinjection and 0.15% ID/g at 30 min postinjection) and showed the labeling of NFTs in an AD brain section by autoradiography assay. These findings indicate the availability of [ ¹¹C]THK-951 for in vivo PET imaging of tau pathology in AD. Copyright

Full text of DOI:10.1002/jlcr.3133

Research Article
Received 3 July 2013,
Revised 31 August 2013,
Accepted 8 September 2013
Published online 16 October 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3133
Synthesis and preliminary evaluation of
2-arylhydroxyquinoline derivatives for
tau imaging
b
b
Tetsuro Tago,a Shozo Furumoto,a,b Nobuyuki Okamura, Ryuichi Harada,
*
Yoichi Ishikawa,a Hiroyuki Arai,c Kazuhiko Yanai,b Ren Iwata,a
and Yukitsuka Kudod
Alzheimer’s disease (AD) is the most common cause of dementia. Senile plaques, consisting of β-amyloid, and neurofibrillary
tangles (NFTs), composed of tau protein, are representative pathological hallmarks of AD. It is believed that the
accumulation of NFTs precedes the onset of clinical symptoms of AD and correlates with the progression of memory
dysfunction. Thus, the use of noninvasive detection techniques including radiolabeled probes and positron emission
tomography (PET) will facilitate early diagnosis or staging of AD. In this study, we synthesized and evaluated novel
hydroxylated 2-arylquinoline derivatives as tau imaging PET probes. The binding affinities of compounds for tau were
evaluated by fluorescent staining of the AD hippocampal section and a competitive binding assay using [18F]THK-523.
THK-951 showed high binding affinity for tau pathology in an AD brain section and K18Δ280K fibrils (Ki = 20.7 nM); thus,
we radiosynthesized a 11C-labeled THK-951 and further studied its potential as a tau PET probe. The [11C]THK-951
demonstrated excellent kinetics in a normal mouse brain (3.23% ID/g at 2 min postinjection and 0.15% ID/g at 30 min
postinjection) and showed the labeling of NFTs in an AD brain section by autoradiography assay. These findings indicate
the availability of [11C]THK-951 for in vivo PET imaging of tau pathology in AD.
Keywords: Alzheimer’s disease; tau; positron emission tomography; imaging
from pathological changes, other quantitatively detectable
biomarkers are desired.
Introduction
Alzheimer’s disease (AD) is a typical neurodegenerative disorder
Tau proteins, which associate with the stabilization of
associated with memory impairment, disorientation, and language
microtubules, are abnormally phosphorylated and form paired
helical filaments (PHFs) in the AD patient’s brain.15,16 PHFs finally
disorder. Because AD shows complex and diverse clinical
symptoms, which overlap with other dementias, its definitive
diagnosis is reliant on postmortem examination rather than clinical
diagnosis. Neuronal atrophy and deposition of senile plaques (SPs)
assemble into NFTs, neuropil threads, and dystrophic neurites.
Neurofibrillary lesions appear in certain brain areas before the onset
of dementia, and autopsy studies indicate a higher association
and neurofibrillary tangles (NFTs) in the brain represent the
between tau pathology levels and cognitive dysfunction relative to
pathological hallmarks of AD.1–3 Consequently, noninvasive
imaging techniques to measure these pathological changes can
be indispensable to the differential diagnosis of AD.
Aβ pathology.17–20 Thus, quantitative imaging of tau burden may
serve to not only monitor the progression of neurodegeneration
but also evaluate the therapeutic effect of longitudinal tau-targeted
AD treatments.
According to amyloid hypothesis, the AD pathogenic
mechanism starts with the abnormal accumulation of SPs
consisting of β-amyloid aggregation in the brain.4 With a view to
presymptomatic diagnosis, investigation into the detection of
lesions using positron emission tomography (PET) or single photon
emission computed tomography has been promoted in this
decade, and many radiolabeled ligands were developed ([18F]
FDDNP,5 [11C]PIB,6 [18F]AV-45,7 [11C]BF-227,8 and [123I/125I]IMPY9).10
These ligands have been successful for the detection of SPs in living
human brains and were useful for confirming the correlation
between PET images and autopsy results.11–13
aCyclotron and Radioisotope Center (CYRIC), Tohoku University, Sendai
980-8578, Japan
bDepartment of Pharmacology, Tohoku University School of Medicine, Sendai
980-8575, Japan
cDepartment of Geriatrics & Gerontology, Institute of Development, Aging and
Cancer, Tohoku University, Sendai 980-8575, Japan
dInnovation of New Biomedical Engineering Center, Tohoku University, Sendai
980-8574, Japan
Although β-amyloid PET imaging studies have high sensitivity
and can detect SPs before clinical symptoms appear, they also
showed that the association between Aβ signals and acuteness
*Correspondence to: Shozo Furumoto, Cyclotron and Radioisotope Center (CYRIC),
Tohoku University, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan.
of symptom was limited.4,14 Therefore, to estimate AD severity E-mail: furumoto@cyric.tohoku.ac.jp
J. Label Compd. Radiopharm 2014, 57 18–24
Copyright © 2013 John Wiley & Sons, Ltd.

T. Tago et al.
Recently, several tau ligands for PET or single photon emission Measurement of logP values
computed tomography imaging have been reported.21,22 Kolb
The octanol/water partition coefficients of the tested compounds were
and colleagues have reported 18F-labeled tau tracers, [18F]T807
and [18F]T808, and these tracers showed promising results in both
in vitro and in vivo studies.23–25 We have developed 2-arylquinoline
derivatives with a preference to bind tau lesions rather than Aβ as
previously reported.26,27 Through the structural optimization, we
recently reported 18F-labeled derivatives [18F]THK-5105 and [18F]
THK-5117.28 They appear to have more preferable properties as a
PET tau imaging radiotracer compared with [18F]THK-523.27 The
preliminary clinical studies using these tracers are currently
in progress.
estimated by HPLC method of the OECD (Organisation for Economic
Co-operation and Development) guideline29 according to the slightly
modified procedures we recently reported.28
In vitro fluorescent binding assay with tau fibrils
Preparation of K18Δ280K-tau fibrils
Recombinant K18Δ280K-tau was custom generated by Life Technologies
Japan Ltd. (Tokyo, Japan). K18Δ280K-tau fibrils were prepared as described
previously.27,30 Briefly, K18Δ280K-tau solution was diluted with phosphate
buffered saline (pH 7.4) to a final concentration of 20μM. The solution was
incubated at 37 °C for a day with agitation (1000rpm). Fibril formation was
confirmed by thioflavin-S fluorescence spectroscopy.
Lipophilic ligands display higher nonspecific binding; therefore,
lower lipophilicity is important for ideal brain kinetics as brain
imaging PET tracers. Low nonspecific binding contributes to high
signal-to-background ratio and facilitates the visualization of lower
burden of tau pathology. The aim of the present study was to
develop a 11C-labeled tau probe with fast clearance from
nonspecific regions. We developed novel hydroxylated quinoline
derivatives (Figure 1) with lower lipophilicity and evaluated their
potential as tau probes for PET.
In vitro fluorescence binding assay
Synthetic K18Δ280K-tau fibrils (200 nM) were incubated with 3 μM (THK-
951 and THK-5272) or 5 μM (THK-953 and THK-5273) compounds at room
temperature for 1 h. The fluorescence intensity was measured at an
excitation wavelength appropriate for each compound. The incubations
were carried out on Nunc 96 MicroWell™ black plates (Nalge Nunc Int.,
Rochester, NY, USA), using 200 μL of phosphate buffer saline (pH 7.4) as
the reaction mixture. All experiments were conducted in triplicate.
Experimental
Synthesis of 2-arylquinoline derivatives
In vitro neuropathological staining using brain sections
Methods for the synthesis and characterization data of 2-arylquinoline
derivatives (shown in Schemes 1 and 2) are described in detail in the
Supporting Information.
The staining properties of the tested compounds were examined using
postmortem brain tissues from an autopsy-confirmed case of AD
(82-year-old woman). The experiments were performed in accordance with
the regulations of the ethics committee of Tohoku University. A series of
8 μm thick sections from paraffin-embedded blocks of hippocampus were
used for staining. According to the method reported previously,28,31 the
sections were immunostained with anti-tau antibody (AT8; Innogenetics,
Ghent, Belgium). Stained images were examined using an Olympus (Tokyo,
Japan) BX51 microscope equipped with blue-violet (BP, 400–440; DM, 455;
BA, 475) and blue filters (BP, 460–490; DM, 500; BA, 520IF).
Radiolabeling of [11C]THK-951
[
11C]Methyl iodide was prepared from [11C]carbon dioxide produced with
a Cypris HM12 cyclotron (Sumitomo Heavy Industries Inc., Tokyo, Japan)
by a catalytic iodination reaction via [11C]methane (MeI MicroLab, GE
Healthcare, Waukesha, WI, USA). [11C]Methyl triflate ([11C]MeOTf) used
for radiolabeling was prepared from [11C]methyl iodide by passing
through a silver triflate-graphpac (Sigma-Aldrich, St Louis, USA) gas
chromatography column heated at 200 °C. The compound 15 indicated
in Scheme 2 was used as a radiolabeling precursor. The precursor
(1.0 mg) and tBuOK (1.0 mg) were dissolved in dimethyl sulfoxide (DMSO)
(400 μL). [11C]MeOTf was bubbled through the solution, followed by
heating at 110 °C for 1 min. Then, 2 N HCl (1 mL) was added to the
solution, and it was heated at an additional 10 min at the same
temperature. The solution was neutralized with 4 N NaOH (0.5 mL), and
the crude mixture was purified with semi-preparative HPLC using an
Inertsil® ODS-3 (GL Sciences Inc., Tokyo, Japan) eluted with acetonitrile/
20 mM NaH2PO4 (pH 4.8) (40/60, flow rate: 5.0 mL/min). The fraction
containing [11C]THK-951 was mixed with water (20 mL) and 25% ascorbic
acid (0.5 mL) and passed through an activated Sep-Pak tC18 Plus (Waters,
Milford, MA, USA). The Sep-Pak cartridge was washed with water, and
then, [11C]THK-951 was eluted with ethanol. The ethanol solution was
used for the in vitro autoradiography (ARG) assay. For the ex vivo
biodistribution study, ethanol was removed by evaporation, and the
product was solubilized into saline with polysorbate 80 (<0.1%).
Competitive binding assay with [18F]THK-523
[18F]THK-523 was synthesized as described previously.27 The product was
formulated with DMSO. The specific activity of [18F]THK-523 was 121.7 GBq/
μmol (end of synthesis), and the radiochemical purity was >99%.
Competitive binding assay was performed as described previously28 with
[18F]THK-523 (final concentration: 4.1 nM) used as a radioligand of the assay.
Autoradiography of [11C]THK-951
The hippocampal brain sections from a case of AD (82-year-old woman)
were used for in vitro ARG. Deparaffinized sections were treated with a saline
solution of [11C]THK-951 (25% ethanol, 0.4 MBq/mL) or [18F]THK-523 (25%
DMSO, 0.6 MBq/mL) for 10 min at room temperature. The sections were
washed briefly with H2O and 50% ethanol. After drying, the labeled sections
were exposed to a BAS-III imaging plate (Fuji Film, Tokyo, Japan) for 3 h ([11C]
THK-951) or overnight ([18F]THK-523). Autoradiographic images were
obtained using the BAS-5000 phosphor imaging instrument (Fuji Film).
Adjacent brain sections were immunostained with anti-tau antibody (AT8;
Innogenetics) and anti-Aβ antibody (6F/3D; Dako, Glostrup, Denmark).
Ex vivo biodistribution assay of [11C]THK-951 in normal mice
Ex vivo biodistribution assay was performed as described previously.28
Briefly, [11C]THK-951 (saline solution: 1.8–1.9 MBq/200 μL) was administered
to male ICR mice (average weight: 29 g) via the tail vein. The mice were
anesthetized with isoflurane and sacrificed by decapitation at 2, 10, 30, 60,
and 90 min postinjection. The brain, blood, and other organs were excised.
Figure 1. Chemical structure of hydroxyquinoline derivatives. Compounds in this
study include the following: X = CH, N; R1 = NH2, NHCH3; R2 = 6-OH, 7-OH.
J. Label Compd. Radiopharm 2014, 57 18–24
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

T. Tago et al.
Scheme 1. Synthesis of hydroxyquinoline derivatives (refer to Supporting Information). A: Synthesis of THK-951. (a) KHMOS, THF. (b)BBr3, CH2Cl2. B: Synthesis of THK-953.
(c) Na2CO3, Pd(PPh3)4, DME/H2O. (d) CH3I, NaH, DMF. (e) 1. TBAF, THF, 2. TFA, CH2Cl2. C: Synthesis of THK-5272 and THK-5273. (f) Na2CO3, Pd(PPh3)4, DME/H2O/
EtOH. (g) NH2CH3, MeOH.
Scheme 2. Synthesis of the precursor (refer to Supporting Information) and radiosynthesis of [11C]THK-951. (a) 1. KHMOS, THF, 2. concentrated HCl, EtOH. (b) 1. BBr3,
CH2Cl2, 2. (Boc)2O, NaHCO3, THF/H2O, 3. aqueous KOH, MeOH. (c) TBDMSCl, imidazole, DMF. (d) 1. [11C]CH3OTf, tBuOK, DMSO, 2. 2 N HCl.
After the organs were weighed, their radioactivity was counted with an
automatic γ-counter. The percentage injected dose per gram (%ID/g) was
calculated by comparing tissue counts to tissue weight. Each % ID/g value
is an average of four separate experiments.
probe candidates expecting they would provide appropriate
brain kinetics.
The synthetic procedures of THK-951, THK-953, THK-5272, and
THK-5273 are shown in Scheme 1 (details of the synthesis
are described in the Supporting Information). THK-951 is a 2-pheny
lquinolin-7-ol derivative synthesized through a condensation
reaction of two phenyl rings (Scheme 1A). The cyclocondensation
Results and discussion
Pittsburgh compound B, a representative β-amyloid probe for of benzaldehyde (1) and acetophenone (2) in the presence of
PET, is a 6-hydroxy substituted derivative of 2-arylbenzothiazole. potassium bis(trimethylsilyl)amide gave 3. THK-951 (4) was obtained
Introduction of a hydroxyl group resulted in low lipophilicity of by the deprotection of 3 by boron tribromide.
the probe and rapid clearance of free radioactivity from the
The THK-953 synthesis was outlined in Scheme 1B. 6 was
brain.6 These characteristics are essential features for brain obtained by Suzuki coupling between 7-hydroxyl-protected
imaging probes because nonspecific signals interfere with quinoline triflate (5) and pyridine boronic acid pinacol ester. After
quantitative analysis of PET images. Therefore, we focused the N-methylation (7), deprotection of the TBDMS and Boc groups
on hydroxyl substituted BF-158 derivatives as tau imaging by treating with TBAF and TFA, respectively, gave THK-953 (8).
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 18–24

T. Tago et al.
The 2-chloroquinolin-6-ol (9), which was prepared from quinolin- in competition with [18F]THK-523, a tau imaging probe for PET. The
2,6-diol by selective chlorination, was used for the synthesis of the Ki values of THK-951, THK-953, THK-5272, and THK-5273 were
6-hydroxy substituted 2-arylquinoline derivatives 10 (THK-5272) 20.7, 110.4, 36.1, and 30.4, respectively (Table 1). THK-951 had the
and 11 via Suzuki coupling reaction (Scheme 1C). THK-5273 (12) highest affinity for tau fibrils from the hydroxyquinoline derivatives
was obtained by
methylamine from 11.
a
nucleophilic substitution reaction with tested. In contrast with the image obtained for its fluorescent
staining, THK-5273 showed comparable affinity with THK-5272.
To estimate the blood–brain barrier permeability of The weak staining of THK-5273 was probably caused by its high
hydroxyquinoline derivatives, logP values were measured by an sensitivity to the washing technique used or the absence of a
HPLC method and found to range from 0.5 to 1.3 (Table 1). These fluorescent hyperchromic effect compared with THK-5272.
data indicated that the hydroxyquinoline derivatives have sufficient
In this study, the quinoline derivatives with higher lipophilicities
had lower Ki values. For Aβ ligands, it is considered that the
lipophilicity to penetrate the blood–brain barrier.27
To evaluate the tau fibril binding characteristics of the synthesized lipophilicity of the ligand correlates with the degree of nonspecific
hydroxyquinoline derivatives, we carried out fluorescence analyses binding but is not correlated with its binding affinity for Aβ
(Figure 2). Fluorescence of the dye thioflavin-T or thioflavin-S fibrils.6,33 It is possible that the lipophilicity of a compound has a
increases when they bind to amyloid fibrils including fibrils from considerable effect on its binding affinity for the cross β-sheet
recombinant tau constructs.30,32 Similar to these dyes, BF-158 structures of tau; however, more structure–activity related studies
derivatives show enhanced fluorescence intensities upon binding are needed to understand this further.
to a cross β-sheet structure. The results from the fluorescence
Following the results of the fluorescence assays and
binding assays using recombinant K18Δ280K-tau fibrils indicated competitive binding assay, we selected THK-951 for additional
that the fluorescence enhancement of THK-951 and THK-5272 was studies using a radiolabeled ligand, with consideration of the
observed in the presence of tau fibrils (Figure 2A, G). In contrast, availability of THK-951 for a tau PET probe. Radiolabeling of
the pyridine-containing derivatives, THK-953 and THK-5273, showed THK-951 was performed using an N-Boc and O-TBDMS protected
little change in fluorescence intensity or, interestingly, slightly precursor (15) and [11C]MeOTf (Scheme 2). [11C]THK-951 was
weakening in the present of tau fibrils (Figure 2D, J). The elaborate obtained in yields of 39% on average with radiochemical purity
verification of the cause of this fluorescence change was not greater than 99% after HPLC purification. The average specific
examined in the present study, but the possibility of a activity of [11C]THK-951 was 83.2 GBq/μmol.
quenching effect caused by the binding conformation of the
compounds is presumed.
The poor brain kinetics of a PET tracer can harm the
quantitative performance of a PET study, so high blood–brain
Subsequently, fluorescent staining using AD hippocampal barrier permeability and rapid clearance from the brain are
sections was performed to examine the in vitro binding affinity essential for brain imaging agents. To evaluate the brain kinetics
of the tested compounds for tau pathology in AD tissues. THK- of [11C]THK-951, we carried out an ex vivo biodistribution study
951 clearly stained NFTs and neuropil threads in the using normal mice (Table 2). [11C]THK-951 showed rapid brain
hippocampal section of an AD brain under a blue filter (Figure 2B, uptake (3.23 0.27% ID/g at 2 min) and smooth washout from
C). THK-5272 also stained NFTs but with a lower contrast image the normal brain (0.11 0.01% ID/g at 60 min), resulting to a high
compared with THK-951 (Figure 2E, F). The neuropathological 2-min-to-60 min ratio (29.4) of brain uptake of [11C]THK-951. This
staining with BF-170, which is a dehydroxylated derivative of uptake ratio in a normal mouse brain is superior to that of [18F]
THK-5272, showed high affinity for NFTs rather than SPs26; thus, THK-523 (1.86), [18F]THK-5105 (9.20), or [18F]THK-5117 (23.1),28
a contribution from the substituted position of the hydroxyl indicating that our structural optimization worked as intended.
group to the affinity for NFTs was indicated. THK-953 (Figure 2H, Compared with these 18F-labeled tracers, the binding affinity of
I) and THK-5273 (Figure 2K, L) are highly fluorescent compounds, [11C]THK-951 for tau aggregates is slightly lower, but the
but very weak images of tau pathology were observed in this superiority in brain kinetics could make up for its weakness in
study. These results correlated with the results from the binding affinity. It is known that the concentrations of tau
fluorescence binding assay using recombinant tau fibrils and aggregates are lower than those of Aβ in an AD brain22;
indicated that the quinoline-pyridine derivatives bind very little therefore, a higher elimination efficiency of the free radiotracer
to regions of tau pathology or have low binding affinity easily from the brain would be more important for tau probes than
removable by a brief wash.
for Aβ probes. The fast clearance of [11C]THK-951 from a normal
For further comparison of the binding affinity of hydroxyquinoline mouse brain is likely caused by its low lipophilicity. The uptake in
derivatives for recombinant tau fibrils, we determined their Ki values the kidney was highest at 2 min postinjection (19.8 % ID/g)
followed by a fast clearance. The uptake in the liver showed a
slow clearance after 10 min postinjection, while the radioactivity
in the small intestine increased with time, indicating the biliary
excretion of radioactive agents.
Table 1. Lipophilicity (logP) and binding affinity (Ki values)
of hydroxyquinoline derivatives to recombinant tau fibrils
Finally, as an assessment of the binding affinity of [11C]
THK-951 for the regions of tau pathology, ARG analysis
using AD hippocampal sections was performed. The ARG
image showed high accumulation of [11C]THK-951 in the
CA1 region of the AD hippocampus (Figure 3A). Distribution
of the tracer signal of [11C]THK-951 correlated well with that
Compound name
X
R1
R2
logP* Ki (nM)
THK-951
THK-953
THK-5272
THK-5273
C
N
C
N
NHCH3 7-OH 1.28
NHCH3 7-OH 0.56
20.7
110.4
36.1
30.4
NH2
6-OH 0.61
of
[
18F]THK-523 (Figure 3B) and tau immunostaining
NHCH3 6-OH 0.90
(Figure 3C). On the other hand, the distribution of the tracer
signal was different from the immunostained Aβ pathology,
which showed spotty distribution in that region (Figure 3D).
*logP values were determined by HPLC method.
J. Label Compd. Radiopharm 2014, 57 18–24
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

T. Tago et al.
Figure 2. Fluorescence spectral analyses and neuropathological staining of THK-951 (A–C), THK-953 (D–F), THK-5272 (G–I), and THK-5273 (J–L). (A, D, G, and J) Fluorescence spectra
of THK-951 (A, λex = 390 nm), THK-953 (D, λex = 390 nm), THK-5272 (G, λex = 390 nm), and THK-5273 (J, λex = 360 nm) in the presence or absence of K18Δ280K-tau fibrils (K18Δ280K-
tau: blue lines; control (ligands only): red lines). (B, C, E, F, H, I, K, and L) Double staining with hydroxyquinoline derivatives and anti-pTau antibody (AT8). (B, E, H, and K) Fluorescent
staining of hippocampal sections from cases of Alzheimer’s disease (THK-951: B; THK-953: E; THK-5272: H; THK-5273: K). (C, F, I, and L) Immunostaining by AT8. Scale bars: 50μm. This
figure is available in color online at wileyonlinelibrary.com/journal/jlcr
assessments of tau burden and other biomarkers of AD (e.g., Aβ
burden, brain metabolism, and clinical findings) are needed to
Table 2. Biodistribution of [11C]THK-951 in ICR mice
understand the accurate pathogenic mechanism or time course
of AD.4 One of the advantages of developing 11C-labeled tau
probes is that tau probes labeled with this short-half-life
radionuclide will facilitate two PET studies in the same day by
using other PET probes such as [18F]FDG or Aβ PET after the
tau imaging.
To develop probes with ideal brain kinetics, we introduced a
hydroxyl group at the 6 or 7 position of quinoline. Lipophilicity
has considerable influences on not only tracer kinetics but also
binding affinity; therefore, lipophilicity needs to be optimized for
tau tracers. In this study, the properties of [11C]THK-951 (logP = 1.28)
were balanced between achieving high binding affinity for tau
fibrils and blood–brain permeability with a fast clearance from
mouse brain. We plan to evaluate other derivatives having different
hydroxylated positions or alkyl amino groups to obtain additional
structure–activity relationship information.
Tissue
Blood
2 min
10 min 30 min 60 min 90 min
2.65
(0.46)
3.23
(0.27)
10.14
(2.66)
19.79
(6.77)
4.72
1.20
0.20
0.12
0.11
(0.48)
0.81
(0.04)
0.15
(0.01)
0.11
(0.02)
0.06
Brain
Liver
(0.13)
10.00
(1.79)
4.10
(0.05)
1.99
(0.45)
0.96
(0.01)
1.10
(0.23)
0.49
(0.01)
0.85
(0.27)
0.35
Kidney
(0.34)
12.22
(3.10)
(0.43)
24.91
(3.17)
(0.11)
29.33
(2.23)
(0.09)
24.64
(1.83)
Small
intestine
(1.44)
Data are expressed as mean of % ID/g (Standard Deviation) (n=4).
These results indicated that [11C]THK-951 was promising in
its ability to bind regions of tau pathology in the AD brain.
In vivo tau imaging will allow the accurate measurement of
the distribution of tau pathology in the AD brain. The combined
A competitive binding assay using a radiolabeled ligand may
ensure the examination of tau binding ligands, which do not
seem to have binding affinity for tau pathology in a fluorescence
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 18–24

T. Tago et al.
Figure 3. Characterization of in vitro binding of [11C]THK-951 to tau pathology by autoradiography and immunohistochemistry. [11C]THK-951 showed high accumulation
in the hippocampal CA1 region (A, filled arrowhead). Similarly, [18F]THK-523, a well-characterized tau probe, also highly accumulated in the CA1 region (B, filled arrowhead).
Immunostaining of adjacent sections demonstrated that numerous tau immunopositive neurofibrillary tangles were observed in CA1 region (C). By contrast, Aβ staining
showed weak immunoreactivity in CA1 region (D).
binding assay using AD brain sections. On the other hand, some
Conflict of Interest
researchers indicated that the PHFs constructed from tau
This study was supported by the research fund from GE Healthcare.
fragments do not fully conform to the structure of PHFs from
an AD brain.22,32 Thus, these in vitro binding assay results should
be considered carefully, and comparison with the results from
References
future clinical studies would facilitate our selection of an
evaluation method.
[1] K. Ogomori, T. Kitamoto, J. Tateishi, Y. Sato, M. Suetsugu, M. Abe, Am.
J. Pathol. 1989, 134, 243–51.
[2] H. Braak, E. Braak, Acta Neuropathol. 1991, 82, 239–59.
[3] P. M. Thompson, K. M. Hayashi, G. de Zubicaray, A. L. Janke, S. E. Rose,
Conclusion
J. Semple, D. Herman, M. S. Hong, S. S. Dittmer, D. M. Doddrell,
A. W. Toga, J. Neurosci. 2003, 23, 994–1005.
We synthesized four hydroxyquinoline derivatives and
performed biological evaluations to estimate their potential as
tau imaging probes. From these results, we selected THK-951
for additional studies and successfully radiosynthesized [11C]
THK-951. [11C]THK-951 showed ideal brain kinetics in normal
mice, and its binding affinity for regions of tau pathology was
confirmed by an in vitro ARG assay using AD hippocampal
sections. Further evaluation of THK-951 and structurally optimized
derivatives will facilitate our understanding about AD tau
pathology in the living body.
[4] C. R. Jack, D. S. Knopman, W. J. Jagust, R. C. Petersen, M. W. Weiner,
P. S. Aisen, L. M. Shaw, P. Vemuri, H. J. Wiste, S. D. Weigand, T. G. Lesnick,
V. S. Pankratz, M. C. Donohue, J. Q. Trojanowski, Lancet Neurol. 2013,
12, 207–16.
[5] E. D. Agdeppa, V. Kepe, J. Liu, S. Flores-torres, N. Satyamurthy,
A. Petric, G. M. Cole, G. W. Small, S. Huang, J. R. Barrio, J. Neurosci.
2001, 21, 1–5.
[6] C. A. Mathis, Y. Wang, D. P. Holt, G. Huang, M. L. Debnath, W. E. Klunk,
J. Med. Chem. 2003, 46, 2740–2754.
[7] S. R. Choi, G. Golding, Z. Zhuang, W. Zhang, N. Lim, F. Hefti,
T. E. Benedum, M. R. Kilbourn, D. Skovronsky, H. F. Kung,
J. Nucl. Med. 2009, 50, 1887–94.
[8] Y. Kudo, N. Okamura, S. Furumoto, M. Tashiro, K. Furukawa,
M. Maruyama, M. Itoh, R. Iwata, K. Yanai, H. Arai, J. Nucl. Med. 2007,
48, 553–561.
Acknowledgements
[9] M.-P. Kung, C. Hou, Z.-P. Zhuang, B. Zhang, D. Skovronsky, J. Q. Trojanowski,
V. M.-Y. Lee, H. F. Kung, Brain res. 2002, 956, 202–10.
[10] M. M. Svedberg, O. Rahman, H. Hall, Nucl. Med. Biol. 2012, 39,
484–501.
[11] V. L. Villemagne, K. E. Pike, D. Darby, P. Maruff, G. Savage,
S. Ng, U. Ackermann, T. F. Cowie, J. Currie, S. G. Chan, G. Jones,
H. Tochon-Danguy, G. O’Keefe, C. L. Masters, C. C. Rowe,
Neuropsychologia 2008, 46, 1688–97.
This study was supported by the Industrial Technology Research
Grant Program of the NEDO (09E51025a) in Japan, the Health
and Labour Sciences Research Grants from the Ministry of
Health, Labour, and Welfare of Japan, Grant-in-Aid for Scientific
Research (B) (23390297), and the ‘Japan Advanced Molecular
Imaging Program (J-AMP)’ of the Ministry of Education, Culture,
Sports, Science, and Technology of Japan.
J. Label Compd. Radiopharm 2014, 57 18–24
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

T. Tago et al.
[12] C. M. Clark, J. A. Schneider, B. J. Bedell, T. G. Beach, W. B. Bilker,
M. A. Mintun, M. J. Pontecorvo, F. Hefti, A. P. Carpenter, M. L. Flitter,
M. J. Krautkramer, H. F. Kung, R. E. Coleman, C. H. Sadowsky, E. M. Reiman,
S. P. Zehntner, D. M. Skovronsky, JAMA 2011, 305, 275–284.
[13] K. Herholz, K. Ebmeier, Lancet Neurol. 2011, 10, 667–70.
[14] V. L. Villemagne, K. E. Pike, G. Chételat, K. A. Ellis, R. S. Mulligan,
P. Bourgeat, U. Ackermann, G. Jones, C. Szoeke, O. Salvado, R. Martins,
G. O’Keefe, C. A. Mathis, W. E. Klunk, D. Ames, C. L. Masters, C. C. Rowe,
Ann. Neurol. 2011, 69, 181–92.
[25] W. Zhang, J. Arteaga, D. K. Cashion, G. Chen, U. Gangadharmath,
L. F. Gomez, D. Kasi, C. Lam, Q. Liang, C. Liu, V. P. Mocharla,
F. Mu, A. Sinha, A. K. Szardenings, E. Wang, J. C. Walsh, C. Xia,
C. Yu, T. Zhao et al., J. Alzheimer’s Dis. 2012, 31, 601–12.
[26] N. Okamura, T. Suemoto, S. Furumoto, M. Suzuki, H. Shimadzu,
H. Akatsu, T. Yamamoto, H. Fujiwara, M. Nemoto, M. Maruyama,
H. Arai, K. Yanai, T. Sawada, Y. Kudo, J. Neurosci. 2005, 25, 10857–62.
[27] M. T. Fodero-Tavoletti, N. Okamura, S. Furumoto, R. S. Mulligan,
A. R. Connor, C. A. McLean, D. Cao, A. Rigopoulos, G. A. Cartwright,
G. O’Keefe, S. Gong, P. A. Adlard, K. J. Barnham, C. C. Rowe, C. L. Masters,
Y. Kudo, R. Cappai, K. Yanai, V. L. Villemagne, Brain 2011, 134, 1089–100.
[15] V. M. Lee, B. J. Balin, L. Otvos, J. Q. Trojanowski, Science 1991, 251,
675–8.
[16] R. Medeiros, D. Baglietto-Vargas, F. M. LaFerla, CNS Neurosci. Ther. [28] N. Okamura, S. Furumoto, R. Harada, T. Tago, T. Yoshikawa,
2011, 17, 514–24.
M. Fodero-Tavoletti, R. S. Mulligan, V. L. Villemagne, H. Akatsu,
T. Yamamoto, H. Arai, R. Iwata, K. Yanai, Y. Kudo, J. Nucl. Med.
2013, 54, 1420–7.
[17] P. V. Arriagada, J. H. Growdon, E. T. Hedley-Whyte, B. T. Hyman,
Neurology 1992, 42, 631–631.
[18] D. W. Dickson, Neurobiol. Aging 1997, 18, S21–S26.
[19] P. T. Nelson, G. A. Jicha, F. A. Schmitt, H. Liu, D. G. Davis, M. S. Mendiondo,
E. L. Abner, W. R. Markesbery, J. Neuropathol. Exp. Neurol. 2007,
66, 1136–46.
[29] OECD GUIDELINE FOR THE TESTING OF CHEMICALS, Partition coefficient
(n-octanol/water), high performance liquid chromatography (HPLC)
method, doi: 10.1787/9789264069824-en.
[30] S. Barghorn, E. Mandelkow, Biochemistry 2002, 41, 14885–96.
[20] P. T. Nelson, I. Alafuzoff, E. H. Bigio, C. Bouras, H. Braak, N. J. Cairns, [31] N. Okamura, T. Suemoto, H. Shimadzu, M. Suzuki, T. Shiomitsu,
R. J. Castellani, B. J. Crain, P. Davies, K. Del Tredici, C. Duyckaerts,
M. P. Frosch, V. Haroutunian, P. R. Hof, C. M. Hulette, B. T. Hyman,
H. Akatsu, T. Yamamoto, M. Staufenbiel, K. Yanai, H. Arai, H. Sasaki,
Y. Kudo, T. Sawada, J. Neurosci. 2004, 24, 2535–41.
T. Iwatsubo, K. A. Jellinger, G. A. Jicha et al., J. Neuropathol. Exp. [32] M. Biancalana, S. Koide, Biochim. Biophys. Acta 2010, 1804, 1405–12.
Neurol. 2012, 71, 362–81.
[21] L. E. Rojo, J. Alzate-Morales, I. N. Saavedra, P. Davies, R. B. Maccioni,
J. Alzheimers Dis. 2010, 19, 573–89.
[22] V. Villemagne, S. Furumoto, M. T. Fodero-Tavoletti, R. Harada, R. S.
Mulligan, Y. Kudo, C. L. Masters, K. Yanai, C. C. Rowe, N. Okamura,
Future Neurol. 2012, 7, 409–421.
[33] A. Forsberg, A. Juréus, Z. Cselényi, M. Eriksdotter, Y. Freund-Levi,
F. Jeppsson, B.-M. Swahn, J. Sandell, P. Julin, M. Schou, J. Andersson,
P. Johnström, K. Varnäs, C. Halldin, L. Farde, S. Svensson, Eur. J. Nucl.
Med. Mol. Imaging 2013, 40, 580–93.
[23] D. T. Chien, S. Bahri, A. K. Szardenings, J. C. Walsh, F. Mu, M.-Y. Su,
W. R. Shankle, A. Elizarov, H. C. Kolb, J. Alzheimer’s Dis. 2013, 34, 457–68.
[24] C.-F. Xia, J. Arteaga, G. Chen, U. Gangadharmath, L. F. Gomez, D. Kasi,
C. Lam, Q. Liang, C. Liu, V. P. Mocharla, F. Mu, A. Sinha, H. Su,
A. K. Szardenings, J. C. Walsh, E. Wang, C. Yu, W. Zhang, T. Zhao
Supporting information
Additional supporting information may be found in the online
et al., Alzheimer’s Dement. 2013, in press, doi: 10.1016/j.jalz.2012.11.008. version of this article at the publisher’s web site.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 18–24