A Gorge-spanning, High-affinity cholinesterase inhibitor to explore β-amyloid plaques

Article abstract of DOI:10.1039/b909612d

Cholinesterases are involved in the pathological formation of β-amyloid plaques. To investigate this pathohistological hallmark of Alzheimer's disease we prepared a high-affinity, fluorescent cholinesterase inhibitor. Its fluorescence intensity was significantly enhanced upon binding to cholinesterases. Using this probe, brain samples from mice and humans affected by Alzheimer's disease were successfully analyzed for β-amyloid plaques. Unexpectedly, it was discovered, by competition experiments, that the compound binds to amyloid structures, rather than to cholinesterases inside of the plaques.

Full text of DOI:10.1039/b909612d

www.rsc.org/obc
Volume 7 | Number 19 | 7 October 2009 | Pages 3881–4132
ISSN 1477-0520
EMERGING AREA
FULL PAPER
Roger D. Rasberry and Ken D. Shimizu
Molecular playdough:
Paul W. Elsinghorst et al.
A gorge-spanning, high-affinity
cholinesterase inhibitor to explore
β-amyloid plaques
conformationally programmable
molecular receptors based on
restricted rotation

PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
A gorge-spanning, high-affinity cholinesterase inhibitor to explore
b-amyloid plaques†
Paul W. Elsinghorst,a Wolfgang Ha¨rtig,b Simone Goldhammer,b Jens Groscheb,c and Michael Gu¨tschow*a
Received 14th May 2009, Accepted 16th June 2009
First published as an Advance Article on the web 17th July 2009
DOI: 10.1039/b909612d
Cholinesterases are involved in the pathological formation of b-amyloid plaques. To investigate this
pathohistological hallmark of Alzheimer’s disease we prepared a high-affinity, fluorescent cholinesterase
inhibitor. Its fluorescence intensity was significantly enhanced upon binding to cholinesterases. Using
this probe, brain samples from mice and humans affected by Alzheimer’s disease were successfully
analyzed for b-amyloid plaques. Unexpectedly, it was discovered, by competition experiments, that the
compound binds to amyloid structures, rather than to cholinesterases inside of the plaques.
became evident. Further improvements of these methods led
to qualitative and semi-quantitative activity estimates of both
AChE and BChE inside b-amyloid plaques.9,10 Recently devel-
oped immuno-(histochemical) techniques allow for the sensitive
detection of both b-amyloid plaques and the co-localized AChE.6
Further investigations of the AChE and BChE activity orig-
inating from b-amyloid plaques revealed, that the inhibition
Introduction
Degeneration of cholinergic neurons in the brain of Alzheimer
patients has led to the cholinergic hypothesis as an attempt to
describe the pathogenesis of Alzheimer’s disease (AD). Accord-
ingly, first symptomatic treatment of AD was realized by inhibiting
the acetylcholine-degrading enzymes, mainly acetylcholinesterase
(AChE).1 The active site of AChE and butyrylcholinesterase
characteristics of cholinesterases located inside b-amyloid plaques
differ notably from those in unaffected tissue.11,12 While it was
shown that Ab itself does not possess cholinesterase activity,13
it remains to be elucidated how the cholinesterase activity is
changed. Thus, it would be worthwhile to establish a probe, that
could reflect functional changes of b-amyloid-associated AChE.
Valuable fluorescent probes, either gorge-spanning14,15 or single-
binding-site directed16,17 irreversible inhibitors, have been devel-
oped. A competitive inhibitor with gorge-spanning properties
addressing both binding sites in native constitution could be
useful to characterize AChE inside b-amyloid plaques (Chart 1).
We envisaged a fluorescent probe originating from previously
established heterodimeric cholinesterase inhibitors.18,19 The novel
probe was designed as a non-irreversible, mixed-type inhibitor
with a fluorophore located at the PAS, while a linker would span
the gorge to place a tacrine-derived moiety at the active site. Tissues
from an AD case and triple-transgenic mice with age-dependent
b-amyloidosis and t-hyperphosphorylation were used to reveal its
histological properties.20
(BChE) is located at the end of a gorge reaching deep into the
protein. Substrates and ligands are guided into and through the
gorge by a negative surface charge.2 The peripheral anionic site
(PAS) of AChE is characterized by a highly conserved tryptophan
that is commonly found cation–p or p–p interacting with aromatic
ligand moieties. An aromatic p–p stacking at the PAS is a feature
often utilized by bifunctional inhibitors simultaneously targeting
both the active and peripheral binding site.3
Senile plaques composed of hardly soluble b-amyloid (Ab)
peptides and neurofibrillary tangles consisting of hyperphospho-
rylated, microtubule-associated protein t are known histopatho-
logical hallmarks of AD. The formation of b-amyloid by either
a- or b-secretase, followed by g-secretase (presenilin-1) leads to
either non-amyloidogenic p3 or to Ab, which might—in patholog-
ically elevated concentration—accumulate in b-amyloid plaques.4
Silver stains, according to Bielschowsky, Gallyas or Braak, are
staining methods indicating both main hallmarks of AD.5 Congo
red and thioflavins selectively target the b-sheet structures in senile
plaques.6 Once staining techniques recognizing cholinesterase
activity had been developed,7,8 the deposition of AChE in plaques
Results
aPharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn,
An der Immenburg 4, D-53121, Bonn, Germany. E-mail: guetschow@
uni-bonn.de; Fax: +49 228 732567; Tel: +49 228 732317
Synthesis
In accordance with previous reports that revealed the coumarin
ring system as
bPaul Flechsig Institute for Brain Research, Department for Neurophysiology,
University of Leipzig, Jahnallee 59, D-04109, Leipzig, Germany
a promising moiety of ligands for both
AChE21 and Ab,22 the fluorogenic moiety was derived from the
7-aminocoumarin skeleton. In particular, its planar electron-rich
aromatic system was anticipated as a versatile interaction partner
for Trp286,3 which plays the fundamental role in PAS binding.
As depicted in Scheme 1, malonic acid (2) was converted to the
activated bis(2,4,6-trichlorophenyl) ester 3 since alkyl esters turned
out to be too unreactive for the following reaction step. Compound
3 was subsequently condensed with 3-diethylaminophenol (4), in
cInterdisciplinary Center of Clinical Research (IZKF), Faculty of Medicine,
University of Leipzig, Inselstraße 22, D-04103, Leipzig, Germany
† Electronic supplementary information (ESI) available: Syntheses and
elemental analyses of compounds 3, 5, 8, and 10–12, Lineweaver–Burk
analysis, fluorescence experiments, labeling of senile plaques with 13 in
the presence and absence of fasciculin-2, histochemical enzyme activity
assay with different concentrations of 13, NMR spectra of compounds
3, 5, 8, and 10–13, and mass spectrometry analysis of 13. See DOI:
10.1039/b909612d
3940 | Org. Biomol. Chem., 2009, 7, 3940–3946
This journal is
The Royal Society of Chemistry 2009
©

Chart 1 Examples of gorge-spanning, linker-connected cholinesterase inhibitors: a pyrene-labelled phosphonofluoridate14 (left) and a huprine–tacrine
heterodimer3 (right). Our design combines a fluorophore and tacrine via a spacer of similar length.
which 7-hydroxyquinolone formation23 is prevented because of
its diethyl substitution, producing the fluorescent 7-diethylamino-
4-hydroxycoumarin (5).24 The attachment of the fluorophore
was intended to proceed via an enamine formation involving
4-aminophenylacetic acid. It was suitable to hold the tacrine
moiety by a hydrazide bond, while the methylene group added
flexibility to the linker, that would have been missed using the more
common 4-aminobenzoic acid. Thus, 4-nitrophenylacetic acid (9)
was converted to its hydrazide 10 via the corresponding ethyl ester
and reacted with 9-chloro-1,2,3,4-tetrahydroacridine (8), which in
turn was obtained by Friedla¨nder condensation of anthranilic acid
(6) and cyclohexanone (7). The reaction joining the tacrine moiety
to the hydrazide involved the use of sealed reaction vessels. The
resulting nitrophenyl-substituted tacrine derivative 11 was then
reduced to 12 using tin(II) chloride dihydrate, as the application of
Pd/C or Ra/Ni interfered with the hydrazide bond. Applying a
reaction originally developed by Knierzinger and Wolfbeis,24 the
enamine formation between 5 and 12 producing 13 was achieved
using triethyl orthoformate in refluxing glacial acetic acid.
at 500 mM concentration of the corresponding thiocholine ester
substrates. Additionally, Lineweaver–Burk analysis of kinetic
measurements in the presence of different substrate and inhibitor
concentrations was carried out (see S5 of the ESI†). Assuming
a mixed-type inhibition mechanism, the dissociation constants
Kic and Kiu were determined from secondary plots, where Kic
(hAChE: 0.21 0.06 nM, hBChE: 3.5 0.1 nM) describes the
competitive and Kiu (hAChE: 0.25 0.04 nM, hBChE: 14 2 nM)
the uncompetitive contribution to binding of 13 to hAChE or
hBChE.18 The resulting value of a = Kiu/Kic (hAChE: 1.2 0.4,
hBChE: 4.0 0.4) was considered to characterize the interaction of
13 with the active and peripheral binding site. For hAChE, a value
close to 1 suggests a similar affinity of 13 for the free enzyme as for
the enzyme–substrate complexes, i.e. the Michaelis complex and
the acetyl-enzyme. Affinity to these enzyme–substrate complexes
may arise from interactions between the inhibitor and amino
acid residues not involved in substrate binding, e.g. the PAS of
hAChE. Due to their architectural differences, peripheral binding
to hBChE is expected to be less pronounced.3 This is reflected in
the observed value a >1, indicating a higher affinity of 13 for the
free enzyme. For both cholinesterases, the competitive inhibition is
expected to result from the typical cation–p stacking of the tacrine
heterocycle slipping in between the active-site residues Trp86/82
(hAChE/hBChE) and Tyr337 (hAChE) or Ala328 (hBChE). Such
a binding mode has been predicted for several heterobivalent
tacrine derivatives.3,25–27
Cholinesterase inhibition
Compound 13 was found to be a potent inhibitor of both human
acetylcholinesterase (hAChE) and butyrylcholinesterase (hBChE)
in vitro. Half-maximal inhibitory concentration (IC50) values of
280 10 pM for hAChE and 16 1 nM for hBChE were determined
Scheme 1 Synthesis of the fluorescent AChE inhibitor 13: a) POCl3, reflux, 3 h; b) PhMe, reflux, 2 h; c) POCl3, reflux, 2 h; d) SOCl2, EtOH, RT, 1 h;
e) N2H4, EtOH/H2O, 80 ◦C, 1 h; f) EtOH, 140 ◦C, 18 h; g) SnCl2 ¥ 2 H2O, EtOH, 70 ◦C, 2 h; h) HC(OEt)3, AcOH, reflux, 10 min.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3940–3946 | 3941
©

Fig. 1 A 3D fluorescence plot of 13 showing an excitation–emission peak
at lexc = 405 nm and lem = 517 nm.
Fluorescence properties and histochemistry
The fluorescence properties of 13 were investigated by a 3D scan,
recording emission spectra from 200–600 nm while excitation was
subsequently tuned from 200–600 nm. A surface plot for the 275–
600 nm interval revealed an excitation–emission peak with lexc
=
405 nm and lem = 517 nm (Fig. 1). For subsequent histochemical
experiments, this sharp excitation–emission peak in a wavelength
range beyond 400 nm was considered very suitable, as only low
background due to tissue auto-fluorescence was expected.
In addition, the influence of protein binding on the fluorescence
properties of 13 was investigated (Fig. 2, see also S6 of the ESI†). A
strong enhancement in fluorescence intensity was observed when
13 was allowed to bind to an excess of hAChE or hBChE in
buffered solution. Unspecific binding as a cause of this finding
was excluded by control experiments using bovine serum albumin
(BSA). In both cases, hAChE and hBChE, an excess of donepezil,
a highly potent AChE inhibitor (IC50 = 5.7 nM),28 displaced 13
from the gorge, resulting in a loss of fluorescence intensity.
The properties of 13 as a histological dye detecting b-amyloid
plaques were revealed by confocal laser scanning microscopy
(CLSM). Plaques containing 13 were found in brain tissue sections
from triple-transgenic mice (Fig. 3A), carrying, as transgenes,
mutated human b-amyloid precursor protein, presenilin-1 and t,
and from an autoptic case with Alzheimer’s disease (Fig. 3B). In
both tissues, b-amyloid plaques were simultaneously visualized
using a rabbit-anti-b-amyloid and red-fluorescent carbocyanine
3-conjugated goat-anti-rabbit IgG (Fig. 3A¢ and 3B¢). Merged
staining patterns clarified the allocation of 13, color-coded in
green, and red-fluorescent b-amyloid-immunoreactivity within
the same senile plaques (Fig. 3A¢¢ and 3B¢¢). Both application
sequences, either 13 or immunostaining first, were carried out and
no difference was observed. Notably, the human tissue showed
that 13 has a tendency to highlight the core regions of b-amyloid
plaques where AChE may actually be expected to serve as a seed
for Ab aggregation.29–31 It has been reported that not only the entire
enzyme, but also truncated peptides, e.g. sequences corresponding
to the w-loop, can act as structural motifs to promote Ab fibril
formation.29,30,32
Fig. 2 An increase in fluorescence intensity (lexc = 405 nm, lem = 517 nm)
was observed when 13 (20 nM) was allowed to bind to hAChE (top).
Section A shows the fluorescence intensity of 13 added to assay buffer
after 200 s. The experiments shown in sections B and C were performed in
the presence of 180 nM BSA (B) or hAChE (C). Proteins were added after a
blank recording of 100 s, followed by equilibration and addition of 13 after
400 s. The same effect was observed for hBChE (bottom: black = assay
buffer, white = BSA, gray = hAChE, hatched = hBChE). Binding of 13 to
hAChE or hBChE could be suppressed by 4 mM donepezil (i.e. the 10-fold
equipotent concentration), where the fluorescence intensity returned to
buffer level. A decrease in fluorescence intensity, observed when donepezil
was solely applied, can be attributed to impaired auto-fluorescence of both
cholinesterases upon donepezil binding.
ESI†). No significant difference in staining by 13 (6.8 mM) was
observed as a result of the presence or absence of fasciculin-
2 (3.1 mM). A reduction in fluorescence after treatment with
fasciculin-2 would have provided ultimate proof of 13 binding
to AChE inside b-amyloid plaques. The unaltered observation
leads to further questions, e.g. whether fasciculin-2 will penetrate
into the b-amyloid plaque core, and whether fasciculin-2 can
bind to AChE inside the lattice of aggregated Ab. To exclude
molecular size as a question of debate, comparable experiments
were carried out using donepezil hydrochloride (Mr = 416.0),
which occupies the enzyme gorge in a similar fashion as expected
for 13. Again, no attenuation of staining intensity caused by 13
(6.8 mM) was observed in brain samples pretreated with donepezil
(145 mM) as shown in Fig. 4. The same result was observed after
simultaneous application of equipotent concentrations (145 mM
donepezil, 6.8 mM 13) or excess of donepezil (1450 mM, 6.8 mM 13).
To corroborate our initial conclusion, that 13 binds to AChE
inside plaques, staining experiments were subsequently repeated
in the presence and absence of fasciculin-2, a peptidic (Mr = 6500 ,
IC50 = 24 nM),33 high-affinity AChE inhibitor (see S8 of the
3942 | Org. Biomol. Chem., 2009, 7, 3940–3946
This journal is
The Royal Society of Chemistry 2009
©

Fig. 3 Double fluorescence labeling of senile plaques with 6.8 mM 13 and
b-amyloid-immunoreactivity in the hippocampus of triple-transgenic mice
with age-dependent b-amyloidosis (A) and in the area 7 of the neocortex
from a case with AD (B). CLSM demonstrates the fluorescence labeling
with 13 (A and B), whereas A¢ and B¢ display indirect immunofluorescence
of the same tissue based on rabbit-anti-b-amyloid and red-fluorescent
carbocyanine 3-conjugated goat-anti-rabbit IgG. The merged figures A¢¢
and B¢¢ reveal the occurrence of both markers in close vicinity within the
same plaques (scale bar = 100 mm).
Fig. 4 Double fluorescence labeling of senile plaques in the hippocam-
pus from a 16-month-old triple-transgenic mouse with age-dependent
b-amyloidosis. CLSM reveals the combined staining with 6.8 mM 13
(A and B) and Ab-immunodetection of the same tissue based on
rabbit-anti-b-amyloid and red-fluorescent carbocyanine 3-conjugated an-
ti-rabbit IgG (A¢ and B¢) after pretreatment of tissue sections with 145 mM
donepezil (A–A¢¢) or without pretreatment (B–B¢¢). The allocation of both
markers in the same plaques becomes obvious in the merged figures A¢¢
and B¢¢ (scale bar = 25 mm).
of dark copper ferrocyanide, that results from the reduction of
soluble copper ferricyanide by thiocholine.34
Binding of 13 to AChE inside b-amyloid plaques seems therefore
rather unlikely.
A significant inhibition of AChE by 13 was observed in the
medial septum-diagonal band complex of the basal forebrain
and the hippocampal CA3 region of a triple-transgenic mouse,
as well as the apparently identical AChE inhibition by 13 in
the cingulate cortices of triple-transgenic and wild-type mice
(Fig. 5). Binding of 13 to tissue-bound AChE was further affirmed
observing a concentration-dependent inhibition. With increasing
To elucidate whether 13 will bind to tissue-bound AChE at
all, we switched from competition experiments to inhibition
measurements of tissue-bound AChE by 13. Tissue sections
from brain areas rich in cholinergic neurons were prepared
and acetylthiocholine-cleaving AChE activity was investigated. A
reduced AChE activity was detected as a decrease in precipitation
Fig. 5 AChE activity (90 min, 37 ◦C) histochemically detected without (A–D) and after pretreatment with 68 mM 13 (A¢–D¢). Medial septum/diagonal
band complex of the basal forebrain from a triple-transgenic mouse (A and A¢). Hippocampal CA3 region in a triple-transgenic mouse (B and B¢).
Cingulate cortices in a triple-transgenic mouse (C and C¢) and a wild-type mouse (D and D¢) (scale bars: A = 300 mm, B–D = 100 mm).
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3940–3946 | 3943
©

concentrations of 13 (between 68 nM and 68 mM), less AChE
activity was detected in the dorsal striatum of a triple-transgenic
mouse (see S9 of the ESI†).
initially favored the former hypothesis. However, this explana-
tion was ruled out by histochemical competition experiments
with known cholinesterase inhibitors. Neither fasciculin-2, nor
donepezil was capable of suppressing the staining by 13. Never-
theless, separate experiments revealed the ability of 13 to inhibit
AChE in several brain areas. It is therefore a unique feature of 13
that it detects b-amyloid plaques and inhibits tissue-bound AChE.
The presence of three aromatic moieties in 13 may contribute to
the high affinity to b-amyloid structures.36,37
Discussion
To establish a probe, that could reflect the interplay between
b-amyloid plaques and AChE, we designed a fluorescent, dual-
mode inhibitor exploiting the unique enzyme architecture. The
inhibitor incorporates two distinct binding moieties, each of
them directed either toward the peripheral, or the active site of
AChE. A convergent synthesis was devised in which the coumarin
fluorophore 5 was combined with the active-site-directed tacrine
derivative 12 where phenylacetic acid was chosen to add sufficient
flexibility. Moreover, this synthon is of comparable length to the
favored 8–9 methylene groups.18
Detailed kinetic analysis revealed that 13 is a picomolar
inhibitor of hAChE, while being 15-fold less active against
hBChE. Furthermore, the dual-binding mode, anticipated from
the tailored structure was confirmed as the analysis showed a
mixed-type inhibition pattern. An a value of 1.2 0.4 was obtained
for hAChE implying virtually equal affinity of 13 for both the free
and substrate-bound enzyme.
A strong enhancement in fluorescence intensity of 13 was
observed in the presence of the target proteins, hAChE and
hBChE, but not BSA. As hBChE lacks a tryptophan at the gorge
entrance, interaction of the fluorophore with this residue at the
PAS of hAChE cannot satisfactorily explain the effect on both
cholinesterases. While the PAS is accessible to water and rather
hydrophilic, probably the hydrophobic environment offered by a
somewhat deeper part of the gorge accounts for the amplified
fluorescence intensity. Alternatively, at least a portion of 13 might
be accommodated along the gorge in a reversed manner with the
fluorophore orientated to its bottom and the tacrine moiety at
the PAS. An additional binding site for tacrine near the PAS
has been predicted from investigations with homodimeric tacrine
inhibitors.35
Donepezil was considered an adequate agent for competition
experiments as it is a highly potent inhibitor with molecular
weight and physicochemical properties similar to 13. When bound
to AChE, the donepezil molecule spans the entire length of the
gorge interacting through aromatic stacking with the anionic
subsite at the bottom of the gorge as well as the peripheral site.3
We have demonstrated that donepezil, in a 10-fold equipotent
concentration, does successfully compete with 13 for accommo-
dation within the active-site gorge. These experiments provide
additional evidence for the binding mode of 13 within the gorge
and, furthermore, support our conclusions from histochemical
competition experiments.
Conclusions
The formation of b-amyloid plaques during the pathogenesis
of AD is tightly linked to the presence of acetylcholinesterase.
Staining methods for these plaques involve the use of antibodies
or fluorescent dyes. We prepared a gorge-spanning, picomolar
inhibitor of acetylcholinesterase carrying a fluorophore suitable
for histochemical applications. The synthetic approach toward
this inhibitor and its cholinesterase inhibiting properties are
reported. Fluorescence characteristics and viability of this novel
probe for histochemical identification of b-amyloid plaques in
transgenic mice with b-amyloidosis or human fixed tissues from
an AD case are presented. Switching from fixed tissues to in vivo
experiments and further in vitro studies will be the subject of future
investigations. In conclusion, 13 appears to be a valuable novel
fluorescent probe for acetylcholinesterase and b-amyloid plaques.
Experimental
General methods and materials
Melting points were determined on a Boe¨tius melting-point
apparatus, and are uncorrected. Compound 11 was prepared
using the Bu¨chi Glas Uster autoclave ‘TinyClave’. Thin-layer
chromatography was performed on Merck aluminum sheets, silica
gel 60 F254. Preparative column chromatography was performed
on Merck silica gel 60, 70–230 mesh. 13C NMR (125 MHz)
1
and H NMR spectra (500 MHz) were recorded on a Bruker
Avance DRX 500 spectrometer; 13C NMR signals were assigned
on the basis of 13C/1H correlation experiments (HSQC and
HMBC). Mass spectra were measured on a MS-50 A.E.I. (EI,
70 eV) and an API2000 (Applied Biosystems) spectrometer (ESI).
Fluorescence spectra were obtained on a Perkin Elmer LS-55
spectrofluorimeter. Elemental analyses were performed with a
Vario EL apparatus. Fluorescence labeling was analyzed with
a confocal laser scanning microscope LSM 510 from Zeiss
which was also equipped with a 405 nm diode and a 475 nm
long pass filter. Histochemical documentation of AChE activity
was performed with an Axiophot (Zeiss, Germany) equipped
with an AxioCam HRC and the program AxioVision 3.1.2.1.
Fasciculin-2 was purchased from Sigma (Steinheim, Germany),
donepezil hydrochloride from Molekula (Dorset, UK). Recom-
binant hAChE and hBChE from human serum were obtained
from Sigma (Steinheim, Germany), bovine serum albumin (BSA,
fraction V) from AppliChem (Darmstadt, Germany). Rabbit
antiserum directed against b-amyloid was obtained from Schering
(Berlin, Germany), carbocyanine 3-conjugated goat-anti-rabbit
antibodies from Dianova (Hamburg, Germany). Entellan was
obtained from Merck (Darmstadt, Germany).
Inhibitor 13 proved suitable for histochemical applications and
enabled us to examine brain tissues from mice and humans
affected by AD. The imaging of b-amyloid plaques by 13 in
both murine and human tissue was confirmed by allocated red-
fluorescent immunolabeling. It was shown for the first time that
b-amyloid plaques can be successfully identified using a fluores-
cent, high-affinity AChE inhibitor. Two possible explanations can
be given for this observation. Compound 13 might either bind to
cholinesterases inside of the b-amyloid plaques, or to the amyloid
structure itself. Based on the high affinity of 13 to AChE, we
3944 | Org. Biomol. Chem., 2009, 7, 3940–3946
This journal is
The Royal Society of Chemistry 2009
©

Preparation of 2-(4-(((7-(diethylamino)-2,4-dioxo-2H-chromen-
3(4H)-ylidene)methyl)amino)phenyl)-N¢-1,2,3,4-
tetrahydroacridin-9-ylacetohydrazide (13)
Fluorescence measurements
Fluorescence spectra were obtained from a solution of 13 in
methanol at a dilution suitable to meet the sensitivity level of the
fluorimeter. The excitation wavelength was increased stepwise by
1 nm from 200 to 600 nm, while emission spectra were recorded in
the same interval. To minimize the influence of diffuse scatter,
control measurements were carried out using neat methanol.
The emission spectra obtained from these control experiments
were subtracted from those investigating 13 by a custom PERL
script and the resulting data were analyzed using Gnuplot.39 The
scattered data obtained from the emission spectra was converted
into a grid using the Gnuplot dgrid3d algorithm, where each
data point is weighted with respect to its corresponding data
point based on the Euclidian distance with a norm power of 4.
The obtained grid was then analyzed with respect to excitation
and emission maxima. The influence of protein binding on the
fluorescence intensity of 13 was measured in assay buffer in the
To a solution of 7-(diethylamino)-4-hydroxy-2H-chromen-2-
one (5; 2.0 mmol, 0.47 g) in refluxing glacial acetic
acid (10 mL) 2-(4-aminophenyl)-N¢-1,2,3,4-tetrahydroacridin-9-
ylacetohydrazide (12; 2.0 mmol, 0.79 g) and triethyl orthoformate
(4.0 mmol, 0.59 g) in glacial acetic acid (10 mL) was added
dropwise over 5 min. The reaction mixture was refluxed for a
further 5 min and stirred for 2 h. The solvent was evaporated
in vacuo, the crude residue was suspended in water (100 mL) and
the suspension was neutralized. A crude precipitate was filtered
off and the aqueous phase was extracted with ethyl acetate (2 ¥
200 mL). The organic phases and a solution of the precipitate
in ethanol were combined, dried using anhydrous sodium sulfate
and evaporated in vacuo. The remaining residue was recrystallized
from methanol/acetonitrile to obtain 13 (0.41 g, 35%) as a yellow
precipitate, mpt 217 ◦C (decomposition). 1H NMR (DMSO-d6) d
1.12 (t, 6 H, J = 6.9 Hz), 1.71–1.81 (m, 4 H), 2.76 (bs, 2 H), 2.88
(bs, 2 H), 3.43 (q, 4 H, J = 7.2 Hz), 3.47 (s, 2 H), 6.37 (app. s,
1 H), 6.66 (dd, 1 H, J = 1.9, 9.1 Hz), 7.29 (d, 2 H, J = 7.9 Hz),
7.30 (app. s, 1 H), 7.48 (d, 2 H, J = 8.5 Hz), 7.51 (app. t, 1 H,
J = 7.7 Hz), 7.70 (app. s, 1 H), 7.71 (d, 1 H, J = 9.2 Hz), 7.77
(s, 1 H), 8.29 (d, 1 H, J = 7.6 Hz), 8.71 (d, 1 H, J = 10.7 Hz),
8.75 (app. bs, 0.5 H), 10.32 (s, 1 H), 13.47 (d, 0.5 H, J = 13.6 Hz);
13C NMR (DMSO-d6) d 12.48, 22.47, 22.67, 24.93, 33.75, 39.53,
44.25, 96.56, 97.30, 108.04, 108.61, 115.85, 118.50, 119.13, 123.05,
123.60, 127.11, 127.99, 128.40, 130.62, 133.92, 136.92, 146.74,
148.51, 152.71, 153.48, 153.75, 156.76, 158.20, 163.14, 169.65,
179.67. EA found: C, 69.1; H, 6.2, N, 11.6. C35H35N5O4 · H2O
absence and presence of hAChE, hBChE and/or BSA at lexc
=
405 nm and lem = 517 nm. A glass cuvette containing assay buffer
(930 mL, 100 mM sodium phosphate, 100 mM NaCl, pH 7.3) was
used to obtain blank recordings. A 3.66 mM solution (50 mL)
of either hAChE, hBChE or BSA was added after 100 s and
equilibrated for approximately 300 s while measuring fluorescence.
Stock solutions of hAChE, hBChE and BSA were prepared
according to their molecular weight and/or activity taken from
vendor specifications. Subsequently, inhibitor solutions (20 mL of
13, donepezil, or both; in assay buffer with 10% methanol) were
added to obtain the following final concentrations, 13: 20 nM (70 ¥
IC50), donepezil: 4 mM (700 ¥ IC50), methanol: 0.2%. Fluorescence
recordings were continued for another 300 s. Linear parts were
extracted from each section after visual inspection and subjected
to standard statistical analysis (Grubbs’ test, boxplot). For details,
see S6 of the ESI.†
+
requires C, 69.2; H, 6.1, N, 11.5%. ESI-MS (C35H36N5O4 ) 590.3.
For syntheses and elemental analyses of compounds 3, 5, 8, and
10–12, see S2 of the ESI†. For NMR spectra of compounds 3,
5, 8, and 10–13, see S10 of the ESI,† and for mass spectrometry
analysis of 13, see S17 of the ESI.†
Histochemistry
A first series of histochemical experiments was performed with
30 mm-thick frozen sections from a paraformaldehyde-fixed au-
toptic tissue block containing the neocortical area 7 from a case
with verified Alzheimer pathology. The tissue sample was used
with permission and following the guidelines of the Local Ethical
Committee in Leipzig, Germany. A second series of stainings
was carried out with frontal sections containing the hippocampi
from 12-month-old triple-transgenic mice carrying, as transgenes,
mutated human b-amyloid precursor protein, presenilin-1 and
t.20 These sections were made from brains perfused with 4%
paraformaldehyde and 0.1% glutaraldehyde, post-fixed overnight
with 4% paraformaldehyde, cryoprotected by equilibration with
30% sucrose and cut at 30 mm thickness using a freezing
microtome. The use of animals in the present study was approved
by the Animal Care and Use Committee of the University of
Leipzig, Germany. All histochemical procedures were started by
extensively rinsing of the 30 mm-thick free-floating sections with
0.1 M Tris-buffered saline (TBS), pH 7.4. The sections were
then incubated with solutions containing 6.8 mM 13 in TBS
for 90 min. After several rinses with TBS, non-specific binding
sites for subsequently applied immunoreagents were blocked with
5% normal goat serum in TBS containing 0.3% Triton X-100.
Cholinesterase inhibition assay
Cholinesterase inhibition was assayed spectrophotometrically at
412 nm in duplicate experiments at six substrate (125, 250, 500,
750, 1000 and 1250 mM) and four to six inhibitor (AChE: 0,
250, 500, 750 pM; BChE: 0, 2 4, 6, 8, 10 nM) concentrations
◦
at 25 C. Assay buffer was 100 mM sodium phosphate, 100 mM
NaCl, pH 7.3. Stock solutions (~3 U mL-1) of human acetyl-
and butyrylcholinesterase in assay buffer were kept at 0 ◦C.
Appropriate dilutions were prepared immediately before starting
the measurement. Acetyl- or butyrylthiocholine (10 mM) and 5,5¢-
dithio-bis-(2-nitrobenzoic acid) (DTNB) (7 mM) were dissolved
in assay buffer and kept at 0 ◦C. A stock solution of 13 was
prepared in 10 mM hydrochloric acid and subsequently diluted in
assay buffer. 50 mL of the DTNB solution, 10 mL of the inhibitor
solution, and 10 mL of the cholinesterase solution were added to
870 mL of assay buffer and thorou◦ghly mixed in a glass cuvette.
After incubation for 15 min at 25 C, the reaction was initiated
by adding 50 mL of the substrate solution or appropriate dilutions
and followed over 5 min. Kinetic analysis was performed using
the software GraFit.38 For Lineweaver–Burk analysis, see S5 of
the ESI.†
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3940–3946 | 3945
©

The sections were then incubated with rabbit-anti-b-amyloid40
(1 : 250 in the blocking solution) overnight. Following three
washes with TBS, the tissue was processed with carbocyanine
3-conjugated goat-anti-rabbit IgG (20 mg mL-1 in TBS containing
2% BSA) for 1 h. In control experiments, the sections were
incubated with the blocking solution without primary antibodies.
Additionally, sections were first immunostained and then reacted
with 13 as described above. Finally, all sections were washed
with TBS and distilled water, mounted onto glass slides, air-dried
and coverslipped with Entellan. The influence of fasciculin-2 and
donepezil on the binding properties of 13 was investigated applying
free-floating, 30 mm-thick tissue sections from 12–16-month-old
triple transgenic mice, which had been prepared as described
above. Sections were extensively rinsed with 0.1 M TBS, pH 7.4
and incubated with 3.1 mM (130 ¥ IC50)33 fasciculin-2 for 16 h or
145 mM (25000 ¥ IC50)28 donepezil for 90 min. Subsequently, the
sections were incubated with 6.8 mM (25000 ¥ IC50) 13 for 90 min,
followed by Ab-immunolabeling and tissue imaging as described
above. In parallel, consecutive sections were processed in the same
manner, but without fasciculin-2 or donepezil pretreatment. In
addition, competition experiments were carried out as described
above, where equipotent concentrations (145 mM donepezil,
6.8 mM 13) or excess of donepezil (1450 mM donepezil, 6.8 mM
13) were simultaneously used for incubation. For the labeling of
senile plaques with 13 in the presence and absence of fasciculin-2,
see S8 of the ESI.†
3 D. Mun˜oz-Torrero and P. Camps, Curr. Med. Chem., 2006, 13,
399.
4 D. J. Selkoe, Physiol. Rev., 2001, 81, 741.
5 T. Uchihara, Acta Neuropathol., 2007, 113, 483.
6 Z. X. Shen, Med. Hypotheses, 2004, 63, 285.
7 M. A. Gerebtzhoff, Cholinesterases: A histochemical contribution to the
solution of some functional problems, Pergamon Press, New York, 1959,
pp 99–101.
8 M. J. Karnovsky and L. A. Roots, J. Histochem. Cytochem., 1964, 12,
219.
9 M. M. Mesulam and M. A. Mora´n, Ann. Neurol., 1987, 22,
223.
10 M. M. Mesulam, C. Geula and M. A. Mora´n, Ann. Neurol., 1987, 22,
683.
11 C. I. Wright, C. Geula and M. M. Mesulam, Proc. Natl. Acad. Sci.
U. S. A., 1993, 90, 683.
12 N. C. Inestrosa and R. Alarco´n, J. Physiol. Paris, 1998, 92,
341.
13 C. Geula, B. D. Greenberg and M. M. Mesulam, Brain Res., 1994, 644,
327.
14 J. R. Saltmarsh, A. E. Boyd, O. P. Rodriguez, Z. Radic´, P. Taylor and
C. M. Thompson, Bioorg. Med. Chem. Lett., 2000, 10,
1523.
15 X. Huang, B. Lee, G. Johnson, J. Naleway, A. Guzikowski, W. Dai and
Z. Darzynkiewicz, Cell Cycle, 2005, 4, 140.
16 G. V. De Ferrari, W. D. Mallender, N. C. Inestrosa and T. L. Rosenberry,
J. Biol. Chem., 2001, 276, 23282.
17 G. Amitai, Y. Ashani, A. Gafni and I. Silman, Neurochem. Int., 1980,
2, 199.
18 P. W. Elsinghorst, J. S. Cieslik, K. Mohr, C. Tra¨nkle and M. Gu¨tschow,
J. Med. Chem., 2007, 50, 5685.
19 P. W. Elsinghorst, C. M. Gonza´lez Tanarro and M. Gu¨tschow, J. Med.
Chem., 2006, 49, 7540.
20 S. Oddo, A. Caccamo, J. D. Shepherd, M. P. Murphy, T. E. Golde, R.
Kayed, R. Metherate, M. P. Mattson, Y. Akbari and F. M. LaFerla,
Neuron, 2003, 39, 409.
Histochemical enzyme activity assay
21 I. Schalk, L. Ehret-Sabatier, Y. Le Feuvre, S. Bon, J. Massoulie and M.
Goeldner, Mol. Pharmacol., 1995, 48, 1063.
22 Y. P. Auberson, Patent WO2003074519. Chem. Abstr. 2003, 139,
245901.
23 R. L. Atkins and D. E. Bliss, J. Org. Chem., 1978, 43, 1975.
24 A. Knierzinger and O. S. Wolfbeis, J. Heterocycl. Chem., 1980, 17,
225.
Two 15-month-old triple-transgenic mice and two 4-month-old
wild-type mice were perfused with 4% paraformaldehyde without
glutaraldehyde followed by post-fixation, cryoprotection and
sectioning. After rinsing with TBS, 30 mm-thick free-floating
frozen sections were incubated for 90 min with buffered solutions
containing 68 mM, 6.8 mM, 680 nM, 68 nM and 0 nM 13.
Following three rinses with TBS, the sections were applied to
AChE enzyme histochemistry with acetylthiocholine iodide as
substrate at 37 ◦C according to the literature.34 The development of
insoluble brown reaction product was stopped after 30 or 90 min,
respectively. Finally, the sections were washed with TBS, briefly
rinsed with distilled water, mounted onto glass slides, air-dried
and coverslipped with Entellan in toluene. For the histochemical
enzyme activity assay with different concentrations of 13, see S9
of the ESI.†
25 A. Musiał, M. Bajdal and B. Malawska, Curr. Med. Chem., 2007, 14,
2654.
26 M. L. Bolognesi, A. Minarini, M. Rosini, V. Tumiatti and C.
Melichiorre, Mini-Rev. Med. Chem., 2008, 8, 960.
27 D. Mun˜oz-Torrero, Curr. Med. Chem., 2008, 15, 2433.
28 H. Sugimoto, Y. Iimura, Y. Yamanishi and K. Yamatsu, J. Med. Chem.,
1995, 38, 4821.
29 G. V. De Ferrari, M. A. Canales, I. Shin, L. M. Weiner, I. Silman and
N. C. Inestrosa, Biochemistry, 2001, 40, 10447.
30 N. C. Inestrosa, M. C. Dinamarca and A. Alvarez, FEBS J., 2008, 275,
625.
31 Y. Mimori, S. Nakamura and M. Yukawa, Behav. Brain Res., 1997, 83,
25.
32 Y. Bourne, Z. Radic´, G. Sulzenbacher, E. Kim, P. Taylor and P.
Marchot, J. Biol. Chem., 2006, 281, 29256.
33 B. Lockhart, M. Closier, K. Howard, C. Steward and P. Lestage,
Naunyn-Schmiedeberg’s Arch. Pharmacol., 2001, 363, 429.
34 J. Andra¨ and Z. Lojda, Histochemistry, 1986, 84, 575.
35 Y.-P. Pang, P. Quiram, T. Jelacic, F. Hong and S. Brimijoin, J. Biol.
Chem., 1996, 271, 23646.
Acknowledgements
The authors thank Frank M. LaFerla and Salvatore Oddo
(University of California, Irvine, USA) for triple-transgenic mice
with age-dependent b-amyloidosis and t pathology, and Thomas
Arendt (University of Leipzig, Germany) for supplying the human
brain sample. The excellent technical assistance of Ute Bauer,
Renate Jendrek, Ulrich Ga¨rtner and Stephanie Hautmann is
gratefully acknowledged.
36 A. Lockhart, L. Ye, D. B. Judd, A. T. Merritt, P. N. Lowe, J. L.
Morgenstern, G. Hong, A. D. Gee and J. Brown, J. Biol. Chem., 2004,
280, 7677.
37 A. A. Reinke and J. E. Gestwicki, Chem. Biol. Drug Des., 2007, 70,
206.
38 R. J. Leatherbarrow, GraFit Version 5.13, Erithacus Software
Ltd.Horley, UK, 2007.
39 Gnuplot Version 4.2, http://www.gnuplot.info (January 22,
2009).
40 T. Siegemund, B. R. Paulke, H. Schmiedel, N. Bordag, A. Hoffmann,
T. Harkany, H. Tanila, J. Kacza and W. Ha¨rtig, Int. J. Dev. Neurosci.,
2006, 24, 195.
References
1 E. Giacobini and R. E. Becker, J. Alzheimers Dis., 2007, 12, 37.
2 C. E. Felder, S. A. Botti, S. Lifson, I. Silman and J. L. Sussman, J. Mol.
Graphics Modell., 1997, 15, 318.
3946 | Org. Biomol. Chem., 2009, 7, 3940–3946
This journal is
The Royal Society of Chemistry 2009
©