Proflavine derivatives as fluorescent imaging agents of amyloid deposits

Article abstract of DOI:10.1016/j.bmcl.2011.03.010

A series of proflavine derivatives for use to further image Aβ amyloid deposits were synthesized and characterized. Aged 3xTg-AD (23 months old) mice hippocampus sections incubated with these derivatives revealed preferential labeling of amyloid plaques. Furthermore an in vitro binding study showed an inhibitory effect, although moderate, of these compounds on Aβ₄₀ fibril formation. This study highlights the potential of proflavine as a molecular scaffold for designing new Aβ imaging agents, its native fluorescence allowing in vitro neuropathological staining in AD damaged brain sections.

Full text of DOI:10.1016/j.bmcl.2011.03.010

Bioorganic & Medicinal Chemistry Letters 21 (2011) 2203–2206
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Proflavine derivatives as fluorescent imaging agents of amyloid deposits
Dominique Garin a, Fatima Oukhatar b, Andrew B. Mahon c, Andrew C. Try c, Michel Dubois-Dauphin d,
b,
Frank M. Laferla e, Martine Demeunynck b, Marcelle Moulin Sallanon a, , Sabine Chierici
⇑
⇑
a INSERM, Unit U877, J Fourier University, La Tronche, France
b Département de Chimie Moléculaire, UMR-5250 & ICMG FR-2607, CNRS-Université Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France
c Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, NSW 2109, Australia
d Department of Pathology and Immunology, CMU-University of Geneva, Geneva 4, Switzerland
e Institute for Memory Impairments and Neurological Disorders, University of California Irvine, Irvine, CA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of proflavine derivatives for use to further image Ab amyloid deposits were synthesized and char-
acterized. Aged 3xTg-AD (23 months old) mice hippocampus sections incubated with these derivatives
revealed preferential labeling of amyloid plaques. Furthermore an in vitro binding study showed an
inhibitory effect, although moderate, of these compounds on Ab40 fibril formation. This study highlights
the potential of proflavine as a molecular scaffold for designing new Ab imaging agents, its native fluo-
rescence allowing in vitro neuropathological staining in AD damaged brain sections.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 3 February 2011
Revised 1 March 2011
Accepted 2 March 2011
Available online 6 March 2011
Keywords:
Proflavine
Click chemistry
Fluorescent markers
Amyloid deposits imaging
The extracellular deposits of b-amyloid peptide (Ab) in senile
plaques and intracellular neurofibrillary tangles (NTF) are known
histopathological hallmarks of Alzheimer’s disease (AD) in brain
areas essential for cognitive functions. Especially, the accumulation
of Ab fibrils in plaques is believed to be fundamental to the initial
development of the disease and to trigger a cascade of events such
as neurotoxicity, oxidative damage, and inflammation leading to
neuronal dysfunction and death.1 Therefore, molecules that
specifically bind b-amyloid aggregates might give access to AD’s
molecular imaging. Thioflavin T, Congo Red, stilbene and amino-
naphthalene derivatives have been identified as markers and are
under investigation for in vivo b-amyloid plaque imaging by posi-
tron emission tomography (PET).2,3 A [18F]fluoroacridine orange
analogue with affinity for Ab aggregates and amyloid plaques has
also been identified.4 To date, it is the first example of acridine
derivatives developed as a marker although, as in most of the re-
ported markers, its aromatic structure appears suitable to interact
with the b-sheet rich Ab fibrils (fAb).
The intrinsic fluorescence of the acridine portion allows the di-
rect observation of the binding. The azide and alkyne derivatives
1b–d are also suitable ‘clickable derivatives’ to allow a further
radiolabeling for in vivo imaging using click methodology, espe-
cially with [11C]methylazide and [18F]fluoroalkynes, or Staudinger
ligation.5–7
We also evaluated the interaction of these compounds with b-
amyloid aggregates in vitro during the fibril formation process.
Indeed, the acridine moiety of our derivatives may interfere with
this process through hydrophobic and p
-stacking interactions.8 In
hybrids 2a–d, by adding lipoic acid or Tröger’s base, we expect a
synergetic effect and thus a better interaction with b-amyloid
species. Lipoic acid, anti-oxidative agent with diverse pharmaco-
logic properties in AD, is known to inhibit fAb formation,9,10
whilst the concave shape exhibited by Tröger’s base analogues
(Fig. 2) may be expected to interact within the grooves created
by beta-sheets.
All acridine derivatives were prepared from 3-acetylamino-6-
aminoacridine and most of them were obtained by simple precip-
itation (Scheme 1). The substitution of propargyl chloroformiate al-
lowed the preparation of 1c in one reaction step with 75% yield.
Additional steps were required to introduce the amino free, azide
or alkyne spacer arms of the other derivatives 1a, 1b and 1d. Nev-
ertheless, 1b and 1d were obtained in efficient yield of 75% and
65%, respectively. In the case of 1a, the low isolated yield of 30%
can be attributed to loss of materials during the purification steps
by precipitation.
We propose here to expand the search for Ab amyloid markers
to other aminoacridine derivatives. We synthesized proflavine
derivatives 1a–d and 2a–d (Fig. 1) and studied their ability to func-
tion as histological dyes to bind Ab amyloid plaques from trans-
genic mouse in vitro.
⇑
Corresponding authors.
E-mail address: sabine.chierici@ujf-grenoble.fr (S. Chierici).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.03.010

2204
D. Garin et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2203–2206
1a
NH2
N
1d
H
O
2a
S
N
H
N
N
N
H
S
2d
N
O
O
N
N
O
AcHN
O
N
N
H
R
N3 1b
O
1c
2b
N
N
S
O
N
N
NH
S
N
2c
N
O
N
N
O
Figure 1. Structure of proflavine derivatives examined in this study.
The ability of compounds 1a–d and 2a–d to inhibit fAb40 forma-
tion and to destabilize preformed fibrils was studied in vitro using
thioflavin T (ThT) assays. The specific binding of ThT to amyloid fi-
brils is generally followed by spectroscopy, and this is the common
way to identify amyloid structures, and to follow their formation
and disaggregation. The compounds were first co-incubated for
O
N3
4
N
O
N
N
O
N
O
Figure 2. Tröger’s base analogues.
eight days at 10 lM with Ab40 (50 lM) in the presence of ThT
and the fluorescence change of the dye at 485 nm was measured.
Results (Fig. 3) are expressed in percent compared to the fluores-
cence of control assay with Ab40 alone.
The inhibitory activity of all tested acridine derivatives is rela-
tively moderate compared to reference inhibitors, especially curcu-
min. Curcumin is a better inhibitor of fAb formation than lipoic
acid.11 It is also known to bind amyloid plaques in vitro and
in vivo.12,13 Where curcumin showed an inhibition of 80% at
The acridine heterodimers made from lipoic acid or Tröger’s
base 2a–d were prepared from 1a–d by ligation via an amide or tri-
azole bond using copper catalyzed azide–alkyne cycloaddition. The
common procedure using copper(II) sulfate and sodium ascorbate
as reducing agent in tbutanol/water solution proved to be efficient
and the hybrids 2b–d were obtained with good yields, 70–75%,
from the appropriate alkyne or azide derivatives 3 and 4. It should
be noticed that the lipoyl derivative 3 and the Tröger’s base ana-
logue 4 were both produced as racemic mixture. Only a single
enantiomer is shown for the Tröger’s base derivatives.
10 lM similar to those reported in the literature, our derivatives
are typically around 50–60% of inhibition, with the exception of
1c which is found slightly better inhibitor compared to the others.
O
S
1) Pht-C2H4-COCl
DBU, DMF, rt
S
Cl
O
2a
N
H
NH2
PyBOP, DBU, DMF, rt
2) MeNHNH2
MeOH, rt
1a
O
S
N
H
1) ClCO-CH2Cl
reflux
S
3
O
O
2b
N3
2) NaN3, DMF
tBuOH/H2O (1/1), 60°C
AcHN
N
NH2
N
80°C
H
1b
CuSO4, ascorbate
O
O
Cl
2c
2d
N
H
O
Py, rt
1c
4
tBuOH/H2O (1/1), 60°C
CuSO
1) ClCO-CH=CH2
DBU, DMF, rt
4, ascorbate
O
H2N
2)
N
H
N
H
1d
MeOH, 45°C
Scheme 1. Synthesis of proflavine derivatives 1a–d and 2a–d.

D. Garin et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2203–2206
2205
capability to bind or not bind fibrils. Indeed a lot of compounds
bind fibrils in vitro without disaggregating them and it may be
rather advantageous for markers. Moreover, the presence of sev-
eral binding sites for probes, distinct from those of ThT, has been
reported on Ab peptide fibrils.14
2d
2c
2b
2a
1d
1c
1b
1a
The in vitro assembled Ab fibrils are commonly used to screen
the potential of probes to further bind amyloid plaques. However,
we chose to directly investigate the ability of our candidates, 1a–d
and 2a–d, to label amyloid b deposits from AD transgenic mice,
which are more representative of those found within human senile
plaques.
LA
Cur
We used triple-transgenic mice (3xTg-AD)15 harboring preseni-
lin-1 (PS1M146V), amyloid–beta precursor protein (APPswe) and tau
(tauP301L) transgene. Brain frozen sections were prepared and incu-
0
20
40
60
80
100
Figure 3. Inhibition % of Ab40 fibril formation. Lipoic acid (LA) and curcumin (Cur)
are used as references. The data are the results of three independent experiments.
bated for 90 min with 100 lM compounds in buffer solution. Na-
tive fluorescence of acridine derivatives is exploited to reveal
labeling of amyloid deposits, using fluorescein optics. Labeled pla-
ques are confirmed by staining of the adjacent sections with thio-
flavin S, a pathological dye known to bind to Ab plaques in the
brain.16 The results are illustrated in Figure 4 for compound 2d.
Twenty-three month aged 3xTg-AD mice hippocampus sections
incubated with 2d revealed labeling of dense core large plaques
(Fig. 4A and B), while only a slight green auto-fluorescence ponctu-
ated labeling is observed onto untreated 3xTg-AD brain slice in the
region of amyloid deposit (Fig. 4C). No labeling is observed in age
matched wild type mice (Fig. 4D). Finally, DAPI labeling does not
recover the green labeling of amyloid plaques, excluding possible
binding of the proflavine derivatives to DNA waste, which may
be present in amyloid deposits.
As for 2d, all other tested compounds clearly stained endoge-
nous plaques of amyloid in the hippocampus of aged 3xTg-AD mice
but with more or less efficiency. Indeed, 2d with 1d are the most
efficient compounds related to the amount of labeled plaques, 1c
and 2c are less efficient markers while 1a, 2a and 2b present an
intermediate behavior to label amyloid deposits. These results
indicate no relationship between the structure of derivatives and
their labeling efficiency, nor any correlation between their ability
to bind plaques and to inhibit fAb formation (Fig. 3). Moreover,
Moreover no benefit of the Tröger’s base or lipoyl moiety in hybrids
2a–d is revealed as compared with their acridine parents 1a–d. On
the contrary, the acridine core is found to be essential in hybrids
since analogs of Tröger’s base and lipoic acid, compounds 3 and
4, are totally inactive as inhibitor of fAb formation, even at a higher
concentration of 50 lM. Whilst our derivatives are certainly not
among the most potent inhibitors, it should be nevertheless noted
that inhibition occurs at 0.2/1 molar ratio of test compound/Ab40
where other reported inhibitors, especially small peptide inhibi-
tors, require higher excess of Ab.
Regarding the fibril-destabilizing effects, none of our acridine
candidates are able to disassemble preformed fibrils in vitro. In-
deed the fluorescence of ThT remains unchanged after the incuba-
tion of fresh fAb40 at 37 °C with our derivatives even at high
compound/Ab ratio of 1/2, conditions under which curcumin and
lipoic acid are reported to disassemble 90% and 60% of fibrils,
respectively.10,11 This ThT assay also reveals that our derivatives
are not able to inhibit ThT binding to synthetic Ab40 fibrils as it is
the case for the fluoroacridine orange probe BF1084 reported in
the literature. Nevertheless, the ability of a compound to disaggre-
gate fibrils or to displace the ThT probe is not to be related to its
Figure 4. Medial aspect of the dorsal hippocampus, coronal sections from 3xTg-AD mouse brain incubated in the presence (A, B) or not (C) of compound 2d at 100
Wild type mice brain section incubated with compound 2d 100 M. Scale bar in A, C, D = 100 m and in B = 30 m.
lM. (D)
l
l
l

2206
D. Garin et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2203–2206
as already observed during the interaction studies with fAb40 for-
mation, no benefit of hybrid structures is revealed as regards the
labeling ability.
deposits) associated with this article can be found, in the online
version, at doi:10.1016/j.bmcl.2011.03.010.
In conclusion, we showed in this study the ability of some pro-
flavine derivatives to interact in vitro with b-amyloid aggregates as
well as with amyloid deposits in brain sections from AD transgenic
mice. Moreover, this work highlighted the labeling potential of
simple proflavine derivatives such as compound 1d. Compared to
the fluoroacridine BF1084 which is reported as marker of amyloid
plaques, the acridine derivatives tested in this study present simi-
lar labeling patterns. It appears, nevertheless, difficult from this
preliminary study to compare their efficiency with this analogue
since experimental labeling conditions used are different. Further
studies are essential to determine if our candidates retain their
labeling ability in human brain sections of AD patients and
in vivo in AD transgenic mice models. Nevertheless the present
study proves the potential of proflavine as molecular scaffold to
design imaging probes of amyloid deposits, its native fluorescence
being a real benefit for a first screening.
References and notes
1. Rauk, A. Chem. Soc. Rev. 2009, 38, 2698.
2. Cai, L.; Innis, R. B.; Pike, V. W. Curr. Med. Chem. 2007, 14, 19.
3. Nordberg, A. Neurology 2004, 3, 519.
4. Suemoto, T.; Okamura, N.; Shiomitsu, T.; Suzuki, M.; Shimadzu, H.; Akatsu, H.;
Yamamoto, T.; Kudo, Y.; Sawada, T. Neurosci. Res. 2004, 48, 65.
5. Schirrmacher, R.; Lakhrissi, Y.; Jolly, D.; Goodstein, J.; Lucas, P.; Schirrmacher, E.
Tetrahedron Lett. 2008, 49, 4824.
6. Mindt, T. L.; Müller, C.; Stuker, F.; Salaza, J.-F.; Hohn, A.; Mueggler, T.; Rudin,
M.; Schibl, R. Bioconjug. Chem. 2009, 20, 1940.
7. Pretze, M.; Wuest, F.; Peppel, T.; Köckerling, M.; Mamat, C. Tetrahedron Lett.
2010, 51, 6410.
8. Chiti, F.; Dobson, C. M. Annu. Rev. Biochem. 2006, 75, 333.
9. Holmquist, L.; Stuchbury, G.; Berbaum, K.; Muscat, S.; Young, S.; Hager, K.;
Engel, J.; Münch, G. Pharmacol. Ther. 2007, 113, 154.
10. Ono, K.; Hirohata, M.; Yamada, M. Biochem. Biophys. Res. Commun. 2006, 341,
1046.
11. Ono, K.; Hasegawa, K.; Naiki, H.; Yamada, M. J. Neurosci. Res. 2004, 75, 742.
12. Yang, F.; Lim, G. P.; Begum, A. N.; Ubeda, O. J.; Simmons, M. R.; Ambegaokar, S.
S.; Chen, P.; Kayed, R.; Glabe, C. G.; Frautschy, S. A.; Cole, G. M. J. Biol. Chem.
2005, 280, 5892.
13. Garcia-Alloza, M.; Borelli, L. A.; Rozkalne, A.; Hyman, B. T.; Bacskai, B. J. J.
Neurochem. 2007, 102, 1095.
Acknowledgments
14. Ye, L.; Morgenstern, J. L.; Gee, A. D.; Hong, G.; Brown, J.; Lockhart, A. J. Biol.
Chem. 2005, 280, 23599.
15. Oddo, S.; Caccamo, A.; Shepherd, J. D.; Murphy, M. P.; Golde, T. E.; Kayed, R.;
Metherate, R.; Mattson, M. P.; Akbari, Y.; LaFerla, F. M. Neuron 2003, 39, 409.
16. Vassar, P. S.; Culling, C. E. Arch. Pathol. 1959, 68, 487.
We thank the Department of Psychiatry, Psychiatric Neuroim-
aging Unit, Geneva, Switzerland, for helpful support in mouse
breeding.
Supplementary data
Supplementary data (the synthesis of compounds 2a–d and the
in vitro study of their interaction with Ab40 aggregates and amyloid