18F-Labeled Benzyldiamine Derivatives as Novel Flexible Probes for Positron Emission Tomography of Cerebral β-Amyloid Plaques

Article abstract of DOI:10.1021/acs.jmedchem.6b01063

Early noninvasive visualization of cerebral β-amyloid (Aβ) plaques with positron emission tomography (PET) is the most feasible way to diagnose Alzheimer’s disease (AD). In this study, a series of flexible benzyldiamine derivatives (BDA) were proposed for

Full text of DOI:10.1021/acs.jmedchem.6b01063

Subscriber access provided by UNIV TORONTO
Article
18
F-labeled Benzyldiamine Derivatives as Novel Flexible Probes
for Positron Emission Tomography of Cerebral #-amyloid Plaque
Zijing Li, Xuran Zhang, Xiaoyang Zhang, Mengchao Cui, Jie Lu, Xiaodong Pan, and Xianzhong Zhang
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 15 Nov 2016
Downloaded from http://pubs.acs.org on November 15, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
18F-labeled Benzyldiamine Derivatives as Novel Flexible Probes for
6
7
Positron Emission Tomography of Cerebral β-amyloid Plaque
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Zijing Li*, †, ⊥, Xuran Zhang†, ‡, ⊥, Xiaoyang Zhang‡, Mengchao Cui‡, Jie Lu‡, Xiaodong Pan*, §,
∥, and Xianzhong Zhang*, †
†
Center for Molecular Imaging and Translational Medicine, State Key Laboratory of Molecular
Vaccinology and Molecular Diagnostics, School of Public Health, Xiamen University, Xiamen,
Fujian 361102, China.
‡
Key Laboratory of Radiopharmaceuticals, Ministry of Education, College of Chemistry, Beijing
Normal University, Beijing 100875, China.
§
Department of Neurology and Geriatrics, Fujian Institute of Geriatrics, Fujian Medical University
Union Hospital, Fuzhou 350001, China.
∥
Key Laboratory of Brain Aging and Neurodegenerative Disease, Fujian Key Laboratory of
Molecular Neurology, Fujian Medical University, Fuzhou 350001, China.
⊥
These authors contributed equally to the work.
Abstract
Early noninvasive visualization of cerebral βꢀamyloid (Aβ) plaques with positron emission
tomography (PET) is the most feasible way to diagnose Alzheimer’s disease (AD). In this
study, a series of flexible benzyldiamine derivatives (BDA) were proposed to bind to the
aggregated βꢀamyloid 1ꢀ42 (Aβ1ꢀ42) with high adaptability, high binding affinity to Aβ1ꢀ42 (6.8
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 36
1
2
3
4
5
6
7
8
9
± 0.6 nM) and rapid body excretion. Methylthio derivative (12) and ethoxyl derivative (10)
were further labeled with 18F directly on their benzene ring and examined as PET probes for
Aβ plaque imaging. [18F]12 displayed 4.87 ± 0.52 %ID/g initial uptake and prompt washout
from normal brain in biodistribution. MicroPETꢀCT imaging of transgenic mice indicated
sufficient retention but lower white matter uptake in the AD transgenetic mouse brain
compared with commercial [18F]AVꢀ45. Our experimental results provide new hints for
developing targeting ligands with flexible framework as efficient Aβ probes for PET of AD
brain.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: benzyldiamine derivatives; Aβ plaques; Alzheimer’s disease; PET imaging agent;
flexible scaffold;
2
ACS Paragon Plus Environment

Page 3 of 36
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
9
Alzheimer’s disease (AD), a widespread and devastating neurodegenerative disease, afflicts a
large population of patients especially seniors.1 The intricate symptoms of AD are considered
to be irreversible,2 therefore early diagnosis as the foundation of the timely interference is
extremely crucial. However, relying on the indirect examination of subsequent deterioration
of mental or cognitive status, traditional clinical diagnosis of AD does not reflect the early
onset neuropathological characteristics3 Tau hypothesis4 and chronic inflammation
hypothesis5 initialized to study AD at molecular level in the past two decades. Extra cellular
deposition of Aβ is still considered as a major AD brains’ pathological hallmark and the most
significant target for molecular imaging.6 Since the deposition of Aβ is found to be 10ꢀ20
years earlier than the beginning of AD symptoms,7 noninvasive visualization of Aβ plaques
such as radiopharmaceutical imaging and fluorescence imaging, have been extensively
investigated.8 Especially, quantitative positron emission tomography (PET) is still the optimal
noninvasive AD diagnosis modality up to date.9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Grouped by their chemical backbone structures, many series of small molecular probes that
can be sorted into four categories of binding sites have been synthesized and evaluated
(Figure. 1).8a, 10 Most of these probes derive from thioflavinꢀT or stilbene and share the same
binding site with 6ꢀiodoꢀ2ꢀ(4’ꢀdimethylaminoꢀ)phenylꢀimidazo[1,2]pyridine, 1 (IMPY). All
these probes were featured with a rigid conjugated planarity to fit into the hydrophobic
binding
channels.11
For
example,
three
clinical
probes
12
(2ꢀ{3ꢀ[18F]Fluoroꢀ4ꢀ(methylamino)phenyl }ꢀ1,3ꢀbenzothiazolꢀ6ꢀol ([18F]flutemetamol),6,
(E)ꢀ4ꢀ(2ꢀ(6ꢀ(2ꢀ(2ꢀ(2ꢀ([18F]ꢀfluoroethoxy)ethoxy)ethoxy)pyridinꢀ3ꢀyl)vinyl)ꢀNꢀmethyl
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 36
1
2
3
4
5
6
7
8
9
benzenamine, 2 ([18F]AVꢀ45)13 and [11C]4ꢀNꢀMethylaminoꢀ4´ꢀhydroxystilbene ([11C]SBꢀ13)14
followed this philosophy. However, these lipophilic radiotracers labeled with either 18F or 11C,
which showed relatively high nonspecific white matter uptakes probably due to the myelin
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
uptakes, are not capable to demarcate the borders of the gray matter where the plaque binding
signals are recognized as the predominant proof of plaque deposition.15 Thus, new PET
probes with both higher affinity and selectivity for Aβ plaques still require further
development.
Figure 1. Chemical structures of the previously reported PET probes of Aβ and the
benzyldiamine derivatives developed in this work. The exclusive binding sites of each
4
ACS Paragon Plus Environment

Page 5 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
category are illustrated, with not mentioned ones belonging to thioflavin T binding site.
Recently, two benzyloxybenzene derivatives (Figure. 1) were reported as potential
Aβꢀtargeted agents, which are available for either 18F or 123/125I labeling starting from the
corresponding precursors for either singleꢀphoton emission computed tomography (SPECT)
or PET.16 Benzyloxybenzene structure is not a conjugated plane but these probes possess
good affinities around 20 nM,16a opening new avenues of flexible compounds to improve
PET imaging of AD. However, the highest affinities of these flexible benzyloxybenzene
ligands rest between 10 and 100 nM, which still need to be improved for highꢀsensitivity PET
imaging. Furthermore, their lipophilicities of current probes are too high (Log D =
3.62ꢀ3.96)11 to cause potential white matter uptake problem.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Here, we described a series of flexible benzyldiamine derivatives (BDAs) and evaluated them
as Aß probes by binding affinities, binding mechanism, specificity and brain uptake profiles.
18
A novel Fꢀlabeling approach starting from 2ꢀnitroterephthalaldehyde as precursor with two
18
aldehyde groups for biꢀlinking were also developed. This pathway attaching F atom on the
terephthalaldehyde in the first step makes least change on BDA structure to minimize both
the affinity loss by the radioactive tag and the molecular weight. The following reductive
amination reaction provides lower lipophilic imine products to reduce nonspecific binding.
Compared with other labeling pathways, we expect this innovation could help finding better
probes for AD diagnosis and introduce more 18F labeled compounds for PET.
Results and Discussion
Chemistry
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 36
1
2
3
4
5
6
7
8
9
BDAs (6ꢀ13) were synthesized following the route in Scheme 1. Dimethylnitroterephthalate
went through hydrolysis, reduction to the dimethanol form, and further oxidized by
pyridinium chlorochromate (PCC) to yield 2ꢀnitroꢀterephthalaldehyde (4), which was the
precursor as well as the intermediate for the reference compounds. Then the nitro group was
replaced by fluorine through nucleophilic substitution with a yield of ~15%. The reference
compounds were further obtained using oneꢀpot reductive amination reactions using 5 for the
first time with different anilines in varied yields (40ꢀ95%). Yellow Schiff bases were formed
during the process as an intermediate, which were immediately reduced by NaBH(OAc)3
with small amount glacial acetic acid added to accelerate the reductive process. All the crude
products were further purified by silica gel chromatography, and had their structures
charactered by 1H NMR, 13C NMR, 19F NMR, and MS/HRMS.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthetic route of the precursor and reference compounds. Reagents and the
.
conditions: a. 1) CH3OH, NaOH, reflux, 2) BH3 S(CH3)2, THF, 40 °C; b. PCC, CH2Cl2, rt; c.
18ꢀcrownꢀ6, KF, DMF, 140 °C; d. NaBH(OAc)3, AcOH, rt.
6
ACS Paragon Plus Environment

Page 7 of 36
Journal of Medicinal Chemistry
1
2
3
4
In vitro binding studies
5
6
7
8
9
The quantitative affinities of BDA compounds (6ꢀ13) for Aβ1–42 aggregates were examined in
the competitive binding assay. [125I]1 was used as the competing radioligand for the
thioflavin T binding site17. 1 and 2 were screened under equal treatment as comparison, for
example same concentration gradient. All tested compounds inhibited [125I]1 binding in a
competitive manner doseꢀdependently as illustrated in Figure 2A. Although compounds 6ꢀ13
only differ on the substituted groups R1 and R2, these substituent groups dramatically affected
the affinities to Aβ1–42 aggregates and resulted in varied Ki values from 6.8 to 2.4E3 nM.
From the various affinities listed in Figure 2B, we can reach three conclusions. (1) The
methoxyl and especially thiomethyl seem to have a favorable effect on affinity than alkyl
alone. (2) When the para sites of the two symmetrical anilines are substituted by alkyl chain
or alkoxy chain, The affinities would increase with the lengthening of the chain. For example,
11 (pꢀCH2CH3) has about 4ꢀfold higher affinity than 7 (pꢀCH3) while 10 (pꢀOCH2CH3) has
about 3ꢀfold higher affinity than 8 (pꢀOCH3). (3) Hydrophobic substitution on ortho site
instead of para site decrease affinity [Ki of 8 (pꢀOMe) > Ki of 13 (oꢀOMe)] with 13 having
nearly no binding. (4) The modification on the para sites of the two symmetrical anilines with
hydrophobic functional group may decrease the affinity while Ki of nonꢀsubstituted 6 (pꢀH)
exceeded 1 ꢁM. Since the higher binding affinity is the foundation of higher sensitivity,
compounds 10 (Ki = 14.9 ± 5.1 nM) and 12 (Ki = 6.8 ± 0.6 nM) were selected for 18F
labeling.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. (A) Inhibition curves of [125I]1 binding to Aβ1–42 aggregates by 1, 2, 6ꢀ13. (B)
Inhibition constants (Ki, nM) of BDAs. Results are given as the mean ± SD, n = 3.
Molecular docking
Molecular docking technique18 was applied to explore the underlying mechanism of these
newly proposed Aβ agents. First of all, geometry optimization of compound 10 and 12 in
water phase was executed at the B3LYP/6–31G level.19 As illustrated in Figure 3A, the
optimized geometry of compound 12 was a stairꢀshape flake with the dihedral angles about
92 degree between each aniline plane and the benzyl plane. Although this most favorable
geometry seems twisted, those sigma bonds are dynamically rotating that the molecule keeps
transforming between a flatter stair and a steep one. Then the binding to Aβ1ꢀ42 fibril model
with twoꢀfold symmetry (PDB ID: 2LMO)20 was investigated by molecular docking. The
result indicated that the similar binding pocket with 1 was revealed (Figure 3B-E). We
speculate that when compound 12 rotates itself into a mildly twisted or near flat geometry
8
ACS Paragon Plus Environment

Page 9 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(dihedral angles between each aniline plane and the benzyl plane <31 degree), it will snuggle
into the channels on the sideꢀchain ladders formed by hydrophobic Val18_Phe20.
Distinguished from other rigid molecule cases, the binding of 12 could happen whenever the
dihedral angle is sharper than 31 degree, which may assign a higher probability. The binding
affinity is a measurement of the attraction strength between a receptor and its ligand, and
could be represented by the binding energy mostly determined by the strength of hydrogen
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
bonding, Van der Waals forces and hydrophobic bonds. The lowest calculated binding energy
16a
for 12 is ꢀ6.8 kcal/mol, more negative compared to ꢀ4.95 kcal/mol for 1 to Aβ1ꢀ42
,
suggesting 12 dose fit tightly into the binding pocket with lower binding energy and potential
higher affinity.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 36
1
2
3
4
5
6
7
8
9
Figure 3. Computational studies. (A) Geometry optimization of compound 12 in water phase.
(B) Geometry variation of “ableꢀtoꢀbind” compound 12, of which the dihedral angles between
each aniline plane and the benzyl plane <31 degree. (C) An overview of 12 and Aβ1ꢀ42 fiber. (D,
E) Molecular docking results indicated that the nearꢀflat geometric configuration of 12 bound
to the channels formed on sideꢀchain ladders by hydrophobic Val18_Phe20, resembling the
computational binding behavior of 1.21
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Radiolabeling
18
Following the scheme in Figure 4A, the F labeled intermediate 2ꢀfluoroterephthalaldehyde
([18F]5) was prepared at 140 °C within 10 min at an average radiochemical yield 74 ± 5% (n
= 5, decay corrected) after purification by a SepꢀPak plus C18 cartridge (Waters, USA). 5ꢀ10
min heating for the nucleophilic substitution of nitro group was proper due to that the
aldehyde on the benzene might be oxidized overtime by DMSO. Then [18F]5 was eluted with
1.0 mL 1,2ꢀdichloroethane and went through a reductive amination reaction with
corresponding anilines to yield [18F]10 and [18F]12, respectively. Glacial acetic acid was
added to accelerate the reduction process to about 15 min. Radiochemical yields in this step
were 51 ± 6% (n = 5, decay corrected), giving an average total yield of 35 ± 6% (n = 5, decay
corrected, at synthesis end). Radiochemical purity after semiꢀpreparation high performance
liquid chromatography (HPLC) purification was more than 95%. The identities of [18F]10 and
[18F]12 were performed by coꢀinjection of the nonradioactive standard compounds (Figure
4B), with time differences that depends on dead volumes between detectors (~ 1.0 min).
Specific activities were calculated to be 49 ± 8 GBq/ꢁmol (n = 5, at the end of synthesis),
comparable with other 18F labeled probes for plaques.
10
ACS Paragon Plus Environment

Page 11 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Radiosynthesis. (A) Radiosynthetic route. Reagents and conditions: a. K18F, K222,
DMSO, 140 °C, 5ꢀ10 min; b. NaBH(OAc)3, AcOH, rt, 15 min. (B) HPLC profile of
coꢀinjection of 10 and [18F]10, 12 and [18F]12. Analytical HPLC conditions: gradient elution
(0ꢀ5 min, 40% B, 5ꢀ10 min, 40ꢀ80% B, 10ꢀ20 min, 80% B, 20ꢀ30 min, 80ꢀ40% B. Eluent A
was phosphate buffer, pH = 7.4, and eluent B was CH3CN.) was performed by an analytical
C18 column (Venusil MP C18, Agela Technologies, 5 ꢁm, 4.6 × 250 mm). UV detector, 254
nm.
Octanolꢀwater partition coefficients
The lipoꢀhydro partition coefficients (Log P) of [18F]10 and [18F]12 were measured in a
1ꢀoctanol/PBS (pH = 7.4) system respectively to estimate their blood brain barrier (BBB)
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 36
1
2
3
4
5
6
7
8
9
penetrating abilities. 1.12 ± 0.09 for [18F]10 and 0.94 ± 0.21 for [18F]12 fall in our ideal range
which is more hydrophilic than traditional Aβ probes (Log P = 2ꢀ3) to allow both sufficient
penetration of BBB and reduced nonꢀspecific binding.22
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biodistribution
After i.v. injection of [18F]10 or [18F]12 in saline containing 2.5ꢀ5% ethanol (v/v) to ICR
(Institute of cancer research) mice, uptakes of radioactivity in organs of interest were
measured and showed in Figure 5. With similar lipophilicity, molecular weights and
structures, the initial uptakes of [18F]10 and [18F]12 at 2 min were 2.10 ± 0.37 %ID/g and
4.87 ± 0.52 %ID/g, respectively. Initial brain uptake of [18F]12 is higher compared with
[18F]10 and is more suitable as a brain imaging probe (Table 1).The brain2min/brain60min ratios
were 5.38 for [18F]10, and 4.38 for [18F]12, indicating similar clearance speed acceptable for
cerebral PET imaging. Bone uptakes of both tracers decreased from 2 to 60 min showing that
there was no significant defluorination in vivo. However, other metabolic natures between
[18F]10 and [18F]12 are distinct from each other. The blood and kidney uptakes of [18F]10
were low and decreasing slowly over time. The lung and liver uptakes both summited at 10
min and remained remarkable until 60 min postꢀinjection. We speculate that hydrophobic
aggregation of [18F]10 might have happened in lung. Elimination was mostly in liver and
partially in stomach and intestinal for [18F]10, because remarkable accumulation of
radioactivity were also detected. In contrast, [18F]12 exhibited no significant concentration in
the liver and stomach, but a quite notable uptake at kidney at 2 min followed by rapid
elimination and continuous intense intestine uptake greater than 4 %ID/g.
12
ACS Paragon Plus Environment

Page 13 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Biodistribution histogram of [18F]10 and [18F]12 in male ICR mice. Expressed as
averages for 5 mice ± standard deviation in % injected dose per gram (%ID/g).
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 36
1
2
3
4
5
6
Table 1. Brain and blood uptakes of [18F]10 and [18F]12 in male ICR micea.
7
8
9
Probe Selected tissue
2 min
10 min
30 min
60 min
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[18F]10
1.46 ± 0.33 0.81 ± 0.09 0.76 ± 0.03 0.97 ± 0.04
2.10 ± 0.37 1.38 ± 0.27 0.68 ± 0.08 0.39 ± 0.04
6.22 ± 1.74 2.44 ± 0.58 0.87 ± 0.21 0.59 ± 0.30
4.87 ± 0.52 2.58 ± 0.10 2.25 ± 0.86 1.11 ± 0.24
Blood
[18F]10
Brain
[18F]12
Blood
[18F]12
Brain
a Expressed as %ID/g, n = 5.
In vitro and in vivo stabilities of [18F]12
After incubating for 60 min in saline or mouse plasma at 37 °C, much more than 95%
radioactivity was identified by HPLC as intact compound [18F]12 (Figure S25 A, B). More
importantly as an in vivo imaging agent for brain, metabolism of [18F]12 in blood and brain of
normal ICR mice was further evaluated by HPLC analysis obtained 60 min postꢀinjection
(Figure S25 C, D). [18F]12 displayed moderate stability in the blood and brain, with 75.9 ±
3.5 % and 62.1 ± 5.3 % of the original probe after 60 min, respectively.
5×FAD mice model
Aβ plaques, gliosis, synaptic markers reduction, and neuron loss have been reported on this
five time FAD (5×FAD) transgenic mice model with a total of five familial Alzheimer’s
disease (FAD) mutations coꢀexpressed.23 These mice were supposed to produce cerebral Aβ
plaques as early as two months old and become severer when older.24 Immunofluorescence
staining with mouse antiꢀAβ monoclonal antibody 6E10 (Covance Antibody, 1:20000) was
14
ACS Paragon Plus Environment

Page 15 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conducted to confirm its existence in the cortex on 5ꢀ11 months models. Compared with the
wild type mouse cortex, typical Aβ plaques with diffused boundary were clearly stained
(Figure 6).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Immunofluorescence staining of Aβ on a brain section of a 5×FAD mouse (male,7
month) with 6E10 antibody. (A, B) Photomicrographs in the cortex field. (C, D) Healthy
control.
MicroPETꢀCT imaging
5×FAD mice (male, 5ꢀ7 month old) and same age male wildꢀtype mice (control) were
employed in microPETꢀCT imaging (n = 3) with [18F]12, [18F]2 and [18F]FDG.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 36
1
2
3
4
5
6
7
8
9
Anesthetization was conducted with 2.5% isoflurane during preparation and 1.5% during
scans. 5 min static wholeꢀbody microPET scans were acquired by an Siemens Inveon
microPETꢀCT (Siemens, Germany) 30 min after intravenous injection of [18F]12 (100 ꢁL,
1.85 ± 0.50 MBq), [18F]2 (100 ꢁL, 1.85 ± 0.50 MBq), [18F]FDG (100 ꢁL, 3.7 ± 0.5 MBq),
respectively. Reconstruction of PET images was performed with threeꢀdimensional ordered
subset expectation maximization (2D OSEM) algorithm. For comparison, regions of interest
(ROIs) were drawn on the brain, cerebellum, bone etc. according to the CT templates and
then added to the coꢀregistered PET images. MicroPETꢀCT images implied that [18F]12 can
differentiate AD and normal brain 30 min post injection, and able to be washed out from the
normal brain regions quickly (Figure. 7A, B). The imaging result are similar compared with
the images by [18F]2 (Figure. 7C, D) but white matter uptake were lower by [18F]12 as the
white arrows marked in Figure. 7A, with low uptake areas and median fissure more clearly
illustrated. The cerebral uptakes of [18F]FDG showed that less radioactivity detained in the
brain of model brains than control, which indicated weak brain functions to assist the AD
brain diagnosis (Figure 7E, F). No notable bone and muscle uptakes were found during the
sustained scans.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
ACS Paragon Plus Environment

Page 17 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. MicroPETꢀCT imaging of brain. (A, B) Transversal brain PET images of [18F]12
superimposed onto CT templates. (C, D) [18F]2. (E, F) [18F]FDG.
Standardized uptake values (SUV) of respective probes were measured and calculated for the
cerebral and listed in Table 2. The 5×FAD mice cerebral SUV of [18F]12 was 3.46 ± 0.33,
slightly lower than [18F]2. However, since its control mice cerebral SUV was even lower, the
cerebral SUV ratio5×FAD/control of [18F]12 was slightly superior than [18F]2.
Table 2. Cerebral SUVs from PET images.
Probe
[18F]12
[18F]2
[18F]FDG
5×FAD cerebral 3.46 ± 0.33a 3.68 ± 0.46 5.46 ± 0.53
Control cerebral 1.03 ± 0.37 1.35 ± 0.27 6.78 ± 0.28
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 36
1
2
3
4
Ratio5×FAD/control
3.35
2.72
0.80
5
6
7
8
9
a SUVꢀbw = average activity in ROI (Bql/mL) / injected dose (Bql) × subject weight (g), n =
3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
A series of fluorinated BDA derivatives beyond the traditional field of effective structures
were designed, synthesized and screened for PET targeting cerebral Aβ plaques. The
aromatic rings of the BDA compounds were connected with flexible sigma bonds and unable
to form a plenary πꢀconjugated system. Taking advantage of the good tolerance of Aβ, BDA
compounds competed well against 1 with different binding affinities to Aβ1ꢀ42 aggregates.
The K values differ from 6.8 ± 0.6 nM, best ever of this kind as a flexible probe, to ~2E3 nM.
i
Molecular docking predicted excellent binding to Aβ fibers. Consequently, [18F]10 and
[18F]12 with favorable binding affinities (K = 14.9 ± 5.1, and 6.8 ± 0.6 nM, respectively)
i
were reproducibly synthesized in radiochemical yields greater than 30%. In biodistribution
study, the 18Fꢀlabeled ligands displayed satisfactory in vivo uptake profile. Especially, [18F]12
exhibited high brain uptake with 4.87 %ID/g 2 min postꢀinjection, moderate clearance with
brain2 min/brain60 min ratio of 4.38, but no significant deflouronation. Meanwhile, [18F]12
displayed acceptable biostability in mouse blood and plasma (intact form > 95% at 60 min).
Decomposition of [18F]12 in the brain can't be ignored (intact form about 62% at 60 min),
which should be improved in the following studies for Aβ imaging. MicroPETꢀCT images
implied that all [18F]12 can penetrate the BBB to target and produce enough signals for
detection which is comparable with the images of [18F]2. The ratiosAD/control of [18F]12 and
18
ACS Paragon Plus Environment

Page 19 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
[18F]2 during proper imaging window are better than the reported ratioAD/control of
[11C]Pittsburgh compoundꢀB in human.25 Since the imaging data on AD model mice by the
18F labeled benzyloxy derivatives has not been reported, comparison of imaging abilities
between flexible probes has not been available yet. In conclusion, Aβ fiber binding pocket
exhibited good tolerance to BDA ligands distortion. And this innovative finding favors the
new standpoint that the planar rigid molecular with πꢀconjugated system was dispensable,
providing new hints for developing targeting ligands with flexible framework for improving
localization and quantitative PET imaging.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental
General information
Reagents used in the synthesis were purchased from J&K Scientific Ltd. (China) and applied
without further purification. The reactions were monitored with Thin Layer Chromatography
on Silica gel 60 F254 aluminum sheets (Merck, Germany), where compounds were
visualized with a 254 nm UV lamp. Column chromatography purifications were conducted
on 54 ꢀ 74 ꢁm silica gel (Qingdao Haiyang Chemical, China). The melting points of newly
synthesized compounds are detected by a MPA100 automated melting point system
1
(OptiMelt, USA). H NMR, 13C NMR and 19F NMR characterization of our synthesized
compounds dissolved in CDCl3 was performed on a 400 MHz Spectrometer (Avance II,
Bruker, Germany) with tetramethylsilane as an internal standard. Mass spectra were acquired
with the amazon SL Bruker instrument or the Bruker Equire 3000 plus (ESI) instrument.
HPLC analysis and purification were performed on a Thermo Scientific Dionex Ultimate
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 36
1
2
3
4
5
6
7
8
9
3000 system equipped with SPDꢀ20A UV detector (λ = 254 nm) as well as a Bioscan Flow
Count 3200 NaI/PMT γꢀradiation detector. Thermo Scientific Hypersil Gold columns
(250×4.6 mm, 250×10 mm) were employed at a flow rate of about 1.0 mL/min for analysis
and 4.0 mL/min for separation. Mobile phase A, phosphate buffer, pH = 7.4. Phase B, HPLC
grade acetonitrile (Amethyst Chemicals, China). The elution sequence was 40% B (0ꢀ5 min),
40ꢀ80% B (5ꢀ10 min), 80% B (10ꢀ20 min), 80ꢀ40% B (20ꢀ25 min). Radioactivity was
measured on a WIZARD2 2480 automatic gamma counter (PerkinElmer, USA). The
centrifuge we used in Log D determination was Eppendorf Centrifuge 5424 R (Germany). All
mice protocols were approved by Xiamen University Animal Care Association. [18F]Fluoride
and [18F]FDG was obtained from the First Affiliated Hospital of Xiamen University.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(2ꢀNitroꢀ1,4ꢀphenylene)dimethanol (3)
5.0 g dimethylnitroterephthalate was dissolved in 80 mL methanol. 22 mL NaOH (2.0 M)
solution was added to the mixed solution which was afterwards heated to 60 °C for 5 h before
cooling down to the room temperature. Then the reaction mixture was quenched with HCl
(2.0 M) and concentrated under vacuum. The precipitation was filtered and dried to obtain
4.3 g product at a yield of 98%. 2.0 g (9.5 mmol) of the solid product was gently dissolved in
50 mL THF in a 100 mL double neck bottle. A solution of boraneꢀmethyl sulfide complex in
THF (2.0 M, 25 mL, 50 mmol) was added dropwise at 0 °C under nitrogen. Then the reaction
mixture was stirred for 18 h at 40 °C before quenching with 20 mL NaOH aqueous solution
(1.0 M). THF was removed in vacuum and the remaining aqueous solution was extracted
with ethyl acetate (2 × 50 mL). After the organic layers was combined and dried with
20
ACS Paragon Plus Environment

Page 21 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
anhydrous sodium sulfate, it was evaporated under reduced pressure to obtain pure
paleꢀyellow solid (1.6 g, 92%). 1H NMR (400 MHz, CDCl3) δ 4.58 (d, 2H, J = 5.82 Hz); 4.79
(d, 2H, 5.28Hz); 5.46 (t, 1H, J = 5.71 Hz); 5.50 (t, 1H, J = 5.58 Hz); 7.68 (d, 1H, J = 8.29
Hz); 7.67 (d, 1H, J = 8.03 Hz); 7.97 (s, 1H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2ꢀNitroterephthalaldehyde (4)
PCC (6.5 g, 30.4 mmol) was added into a solution of 3 (1.4 g, 7.6 mmol) in 200 mL CH2Cl2
while stirring. The reaction mixture was kept stirred at room temperature for 5 h before a
second portion of PCC (1.6 g, 7.6 mmol) was added and kept stirred further for 2 h. We
concentrated the reaction mixture to 50 mL under vacuum, and loaded it onto a silica gel
column, which was eluted with CH2Cl2, fractions containing the product was collected (TLC:
ethyl acetate/petroleum ether 1:2, Rf 0.5) and evaporated under vacuum to yield a
1
paleꢀyellow solid product (0.8 g, 58%). H NMR (400 MHz, CDCl3): δ 8.07 (d,1H, J = 7.74
Hz); 8.37 (dd, 1H, J1 = 7.87 Hz, J2 = 1.02 Hz); 8.61 (d, 1H, J = 1.43 Hz); 10.17 (s, 1H); 10.31
(s, 1H).
2ꢀFluoroterephthalaldehyde (5)
Anhydrous potassium fluoride (64 mg, 1.1 mmol), 18ꢀcrownꢀ6 (300 mg, 1.1 mmol) were
dissolved in 1.0 mL anhydrous DMF in a 10 mL serum vial (WHEATON, USA). After 5 min
stirring at room temperature, 4 (100 mg, 0.65 mmol) dissolved in 0.2 mL anhydrous DMF
was added into the vial. This mixture was placed into 140 °C oil bath for 7 min. After
quenching by 5 mL water, the mixture was extracted with ethyl acetate (2 × 5 mL). Organic
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 36
1
2
3
4
5
6
7
8
9
layers were collected and dried. Evaporated under vacuum, the resulting residue was purified
by column chromatography on silica gel to yield 5 (12 mg, 14%). TLC rf 0.4 (ethyl
acetate/petroleum ether 1:4). 1H NMR (400 MHz, CDCl3): δ 7.71 (d, 1H, J = 9.90 Hz); 7.81
(d, 1H, J = 7.75 Hz); 8.07 (t, 1H, J = 6.73 Hz); 10.09 (d, 1H, J = 1.76 Hz); 10.46(s, 1H). 19F
NMR (376 MHz, CDCl3): δ ꢀ119.37 (s, 1F).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General procedure for the reductive amination reaction (A).
A mixture of 2ꢀfluoroꢀ1,4ꢀbenzenedialdehyde (1 equivalent), aniline (2 equivalent) and
sodium triacetoxyborohyride (2 equivalent) in 1,2ꢀdichloroethane was stirred for 10 min
before AcOH (1 equivalent) was added. 3 h later, 1.0 M NaOH (1 mL) was added to quench
the reaction. The solution was drawn with ethyl acetate (3 × 3 mL). The organic phase was
collected, dried and concentrated under reduced pressure to produce a pale yellow oil residue.
The crude product was purified through column chromatography to yield a white solid
product.
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))dianiline (6)
This compound was prepared in the same way detailed in general procedure A as a white
1
solid (44.1%, m.p. 105 ± 2 °C). TLC Rf = 0.5 (ethyl acetate/petroleum ether 1:6). H NMR
(400 MHz, CDCl3) δ 4.33 (s, 2H); 4.38 (s, 2H); 4.37 (s, 2H); 6.64 (dd, 4H, J1 = 13.81 Hz, J2 =
19
7.69 Hz); 6.74 (t, 2H, J = 7.33 Hz); 7.11 (m, 2H); 7.19 (m, 4H); 7.35 (t, 1H, 7.70 Hz). F
NMR (376 MHz, CDCl3) δ ꢀ118.76 (s, 1F). 13C NMR (CDCl3, 100 MHz): δ 114.9, 115.9,
116.2, 121.6, 124.9, 125.5, 130.1, 130.5, 141.0, 144.1, 149.2, 162.6, 165.1, 188.7. HRMS
22
ACS Paragon Plus Environment

Page 23 of 36
Journal of Medicinal Chemistry
1
2
3
4
+
calcd for C20H20FN2 , 307.1606, found 307.1608 [M + H]+.
5
6
7
8
9
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀmethylaniline) (7)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This compound was obtained in the same way detailed in general procedure A as white solid
1
(55%, m.p. 112 ± 1 °C). H NMR (400 MHz, CDCl3) δ 2.24 (s, 6H); 4.30 (s, 2H); 4.36 (s,
2H); 6.56 (dd, 4H, J1 = 14.52 Hz, J2 = 8.20 Hz); 6.99 (d, 4H, J = 7.68 Hz); 7.08 (d, 2H, J =
10.45 Hz); 7.33 (t, 1H, J = 7.76 Hz). 19F NMR (376 MHz, CDCl3) δ ꢀ118.76 (s, 1F). HRMS
+
calcd for C22H24FN2 , 335.2, found 334.9 [M + H]+.
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀmethoxyaniline) (8)
This compound was obtained in the same way detailed in general procedure A as a pale
yellow solid (95%, m.p. 116 ± 1 °C). 1H NMR (400 MHz, CDCl3) δ 3.73 (s, 6H); 4.26 (s, 2H);
4.32 (s, 2H); 6.59 (dd, 4H, J1 = 23.69 Hz, J2 = 8.80 Hz); 6.76 (d, 4H, J = 6.89 Hz); 7.08 (m,
2H); 7.32 (t, 1H, J = 7.56 Hz). 13C NMR (CDCl3, 100 MHz): δ 42.6, 48.5, 55.8, 114.3, 114.9,
19
129.7.5, 141.4, 142.0, 152.4, 160.3. F NMR (376 MHz, CDCl3) δ ꢀ118.76 (s, 1F). HRMS
+
calcd for C22H24FN2O2 , 367.1817, found 367.1808 [M + H]+.
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀchloroaniline) (9)
This compound was obtained in the same way detailed in general procedure A as a pale
1
yellow solid (40%, m.p. 109 ± 1 °C). H NMR (400 MHz, CDCl3) δ 4.29 (s, 2H); 4.34 (s,
2H); 6.53 (dd, 4H, J1 = 21.36 Hz, J2 = 8.62 Hz); 7.05ꢀ7.12 (m, 6H); 7.29 (t, 1H, 7.59 Hz). 19F
+
NMR (376 MHz, CDCl3) δ ꢀ118.76 (s, 1F). HRMS calcd for C22H18Cl2FN2 , 375.0826, found
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 36
1
2
3
4
375.0819 [M + H]+.
5
6
7
8
9
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀethoxyaniline) (10)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Similar to above, this compound was prepared in the same way detailed in general procedure
A as pale yellow solid (45%, m.p. 119 ± 1 °C). 1H NMR (400 MHz, CDCl3) δ 1.38 (t, 6H, J =
6.98 Hz); 3.96 (q, 4H, J = 6.95); 4.27 (s, 2H); 4.33 (s, 2H); 6.59 (dd, 4H, J1 = 15.48 Hz, J2 =
8.75 Hz); 6.78 (d, 4H, J = 8.90 Hz); 7.08 (d, 1H, J = 3.59 Hz); 7.11 (s, 1H); 7.34 (t, 1H, J =
+
7.54 Hz); 19F NMR (376 MHz, CDCl3) δ ꢀ118.76 (s, 1F). HRMS calcd for C22H28FN2O2 ,
395.2130, found 395.2123 [M + H]+.
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀethylaniline) (11)
1
Similar to above, this compound was obtained as a white solid (43%, m.p. 119 ± 1 °C). H
NMR (400 MHz, CDCl3) δ 1.20 (t, 6H, J = 7.57Hz); 2.56 (q, 4H, 7.54 Hz); 4.31 (s, 2H); 4.37
(s, 2H); 6.59 (dd, 4H, J1 = 14.24 Hz, J2 = 8.05 Hz); 7.02 (d, 4H, J = 7.92 Hz); 7.08 (d, 1H, J =
3.59 Hz); 7.12 (s, 1H); 7.35 (t, 1H, J = 7.52 Hz). 13C NMR (CDCl3, 100 MHz): δ 15.9, 27.9,
42.0, 47.9, 113.0, 114.2, 122.9, 128.6, 129.7, 133.7, 144.3, 145.8, 160.3. 19F NMR (376 MHz,
+
CDCl3) δ ꢀ118.76 (s, 1F). MS calcd for C22H28FN2 , 363.2, found 362.8 [M + H]+.
N,N'ꢀ((2ꢀFluoroꢀ1,4ꢀphenylene)bis(methylene))bis(4ꢀ(methylthio)aniline) (12)
This compound was obtained in the same way detailed in general procedure A s a redꢀbrown
1
solid (69%, m.p. 121 ± 1 °C). H NMR (400 MHz, CDCl3) δ 2.41 (s, 6H); 4.32 (s, 2H); 4.37
(s, 2H); 6.58 (dd, 4H, J1 = 13.96 Hz, J2 = 8.59 Hz); 7.08 (d, 1H, J = 5.40 Hz); 7.10 (s, 1H);
24
ACS Paragon Plus Environment

Page 25 of 36
Journal of Medicinal Chemistry
1
2
3
4
7.21 (d, 4H, J = 8.12 Hz); 7.31 (t, 1H, J = 8.10 Hz). 19F NMR (376 MHz, CDCl3) δ ꢀ118.76 (s,
5
6
7
+
1F). HRMS calcd for C22H24FN2S2 , 399.1360, found, 399.1353 [M + H]+.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N,N'ꢀ((2ꢀ(Fluoroꢀ18F)ꢀ1,4ꢀphenylene)bis(methylene))bis(2ꢀmethoxyaniline) (13)
Similar to the above, this compound was prepared in the same way detailed in general
1
procedure A as a paleꢀyellow solid (48%, m.p. 116 ± 1 °C). H NMR (400 MHz, CDCl3) δ
1.55 (s, 6H); 4.29 (s, 2H); 4.34 (s, 2H); 6.53 (dd, 4H, J1 = 21.52 Hz, J2 = 8.71 Hz). 13C NMR
(CDCl3, 100 MHz): δ 41.3, 47.3, 55.4, 109.5, 110.1, 114.1, 116.9, 121.3, 122.9, 125.1, 129.6,
19
137.8, 141.2, 146.9, 160.3. F NMR (400 MHz, CDCl3) δ ꢀ118.76 (s, 1F). HRMS calcd for
+
C22H24FN2O2 , 367.1817, found, 367.1809 [M + H]+.
In vitro binding assay
In vitro binding assay was conducted in the same way detailed in a previous literature.26
Molecular docking
Compound 12 was constructed with GaussView 5.0 (Gaussian, Inc., USA) firstly. Geometric
optimizations were performed in the water phase in Gaussian 09 Revision C.01 (Gaussian,
Inc., USA) using B3LYP/6ꢀ31G19 for all atoms. Docking simulations on Aβ fibers (PDB ID:
2LMO20) performed according to previously described methods.16a For the docking
simulations, the CꢀN bonds were kept flexible. The grid size was chosen to occupy the whole
ligandꢀpeptide complex, and the spacing was kept to 1.00 Å, a standard value for Autodock
Vina27. Each docking trial produced 100 poses with the exhaustiveness value of 100.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 36
1
2
3
4
5
6
7
8
Radiolabeling
9
The F radiolabeled intermediate [18F]5 was prepared according to the following procedure.
18Fꢀ was trapped, eluted and dried following a routine procedure described in a previous
literature.26 5.0 mg Nitro terephthalaldehyde dissolved in 0.5 mL anhydrous dimethyl
sulphoxide (DMSO) was added into the serum vial containing the K222ꢀ18Fꢀ complex and kept
heating at 140 °C. After 5 min reaction, the vial was taken out from the oil bath, cooled down
to ambient temperature by running tap water. 8 mL water was added into the mixture, before
this DMSO/water solution was passed through a SepꢀPak plus C18 cartridge (waters, USA).
Washed with additional 5 mL water, and the desired intermediate [18F]5 was eluted by ~1.0
mL of 1,2ꢀdichloroethane into a new serum vial. The appropriate amines (14 or 15, 2 mg) and
NaBH(OAc)3 (5 mg) were added into the eluted solution. Stirred for 2 min at room
temperature, glacial acetate acid (4 ꢁL) was injected in to promote the reductive amination
process. We stirred the mixture at room temperature for another10 min. NaOH aqueous (1.0
M, 100 ꢁL) was added to quench the reaction. Then the radiolabeled product was exacted
with ethyl acetate (1 mL), and dried over NaSO4. The organic solvent was evaporated under
nitrogen gas and the residue was dissolved in HPLC eluent for separation. After HPLC
purification, the radiochemical purity of [18F]10 and [18F]12 was bigger than 95%. The
retention time of [18F]10 and [18F]12 were 16.45 and 17.47 min, respectively. For formulation,
the fluid collected from HPLC (HPLC mobile phase during product collection was 80%
acetonitrile and 20% water) was dried with N2 flow and further dissolved in saline mixed
with 2.5% ([18F]10) or 5% ([18F]12) ethanol (v/v) to make the final concentration to be 18.5
18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
ACS Paragon Plus Environment

Page 27 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MBq/mL. Purity and specific activity were sent to analytical HPLC before experiments in
vitro or in vivo. 0.1 mL of this probe solution was injected into each model mice to deliver
1.85 MBq for both biodistribution and imaging experiments. Our injected doses were all
lower than 2.0 MBq with specific activity determined to be 49 ± 8 GBq/ꢁmol (n = 5, at the
end of synthesis) in saline containing no more than 5% (v/v) ethanol, which met the
requirement of safety for animal in this proofꢀofꢀconcept study.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Partition Coefficient Determination
Partition coefficients (Log P) of [18F]10 and [18F]12 were measured in 1ꢀoctanol and
phosphate buffered saline (PBS), which were preꢀequilibrated with each other. About 10 µCi
of [18F]10/12 in water (0.1 mL) after HPLC purification was added in a plastic centrifuge
tube (GEB, Hangzhou Gene Era Biotech Co., Ltd., China), to which 1.9 mL of PBS and 2 mL
octanol was added followed. The tube was vortexed for 5 min followed by 5 min of
centrifugation. Aliquots of 0.1 mL from each phase were measured in an automatic gamma
counter. Partition coefficient was expressed as Log P. P (equals to D in this study) was
calculated as the formula: D = (Counts in 1ꢀoctanol/Counts in PBS). The measurement was
performed in triplicate.
In vitro and in vivo stability studies
After incubation or in vivo biodistribution of [18F]12, the plasma (the mice were sacrificed by
decapitation for blood collection) or ultrasonicated organ samples were precipitated by
adding 100 µL acetonitrile and centrifuged at 3000 rpm for 5 min. 10 µL supernatants were
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 36
1
2
3
4
analyzed by HPLC respectively.
5
6
7
8
9
Immunofluorescence study
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In vitro binding assay was conducted in the same way detailed in a previous literature.24
Biodistribution
Twenty ICR mice (male, 18ꢀ20 g, 5ꢀ6 weeks old) were allowed water and food ad libitum,
and randomly selected for each group by removing from their cages without selection. The
injected doses were measured by a CRCꢀ25R dose calibrator (CAPINTEC, USA) with the
syringe residue activities deducted. 1.85 ± 0.50 MBq of [18F]10 or [18F]12 in 0.1 mL saline
was injected via the tail vein. 2, 10, 30, 60 min postꢀinjection, mice were sacrificed by
decapitation and interested organs were collected, weighted in preꢀweighed plastic bags. The
gall bladders were not included with the livers, and the pancreases were separated from the
intestines. The whole tails were also measured and deducted to determine the total activities
that entered the circulation. Activities in organs were measured by a WIZARD2 2480
automatic γꢀcounter (Perkin Elmer, USA, ~70% efficiency). 100 ꢁL (same volume as injected)
of 100 times dilution of the injected dose as 1 %ID was counted under the same treatment.
The uptake in organs was expressed as the percentage of the injected dose per gram organ
(%ID/g) calculated as the ratio of count per minute (cpm).
microPETꢀCT imaging with an AD transgenic model
1.85 ± 0.50 MBq [18F]12 in 100 ꢁL saline containing 5% ethonal were injected into 7ꢀmonth
28
ACS Paragon Plus Environment

Page 29 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
old 5×FAD mice (male, 22ꢀ26 g) via the caudal lateral tail vein. 20 min post injection, mice
were anesthetized using isoflurane (delivered as a gas mixed with air, 2.5% isoflurane for
preparation and 1.5% during scan) and placed on the bed of microPETꢀCT machine (Siemens,
Inveon). 30 min post injection, a static scan was acquired for 5 minutes while the animal was
kept warm by a heated pad on the scanner bed. A baseline lowꢀdose CT scan was then
obtained for localization and attenuation correction. PET images were reconstructed by a 2D
OSEM/MAP reconstruction algorithm by the Inveon Acquisition Workplace Software
(Siemens, Germany). [18F]2 (injected dose 1.85 ± 0.50 MBq), [18F]FDG (injected dose 3.7 ±
0.5 MBq) were also injected in the tail vein of 5×FAD or normal mice. Images were
acquisitioned under same treatment for comparison.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgements
This work was funded by National Natural Science Foundation of China (81501534,
21271030, 81471707, 81402648 and 81200991), National Key Basic Research Program of
China (2014CB744503) and Young and Middle
‑
aged Talent Training Key Project in Health
System of Fujian Province (2014
interest.
‑ZQN
‑
ZD
‑11). The authors declare no competing financial
We thank Prof. Huabei Zhang (College of Chemistry, Beijing Normal University) for
assistance in the molecular docking.
Supporting Information Available: characterization of the precursor and reference
compounds, purity of newly synthesized compounds, stability study of [18F]12, standard
curve of UV absorption of compound 12, preparation and quality control data of [18F]AVꢀ45,
29
ACS Paragon Plus Environment