Multi-target-directed coumarin derivatives: hAChE and BACE1 inhibitors as potential anti-Alzheimer compounds

Article abstract of DOI:10.1016/j.bmcl.2007.09.100

The complex etiology of Alzheimer's disease (AD) prompts scientists to develop multifunctional compounds to combat causes and symptoms of such neurodegeneration. To this aim we designed, synthesized, and tested a series of compounds by introducing halophenylalkylamidic functions on the scaffold of AP2238, which is a dual binding site acetylcholinesterase inhibitor. The inhibitory activity was successfully extended to the beta-site amyloid precursor protein cleavage enzyme, leading to the discovery of a potent inhibitor of this enzyme (3) and affording multifunctional compounds (2, 6, 8) for the treatment of AD.

Full text of DOI:10.1016/j.bmcl.2007.09.100

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 423–426
Multi-target-directed coumarin derivatives: hAChE and BACE1
inhibitors as potential anti-Alzheimer compounds
Lorna Piazzi,* Andrea Cavalli, Francesco Colizzi, Federica Belluti, Manuela Bartolini,
Francesca Mancini, Maurizio Recanatini, Vincenza Andrisano and Angela Rampa
Department of Pharmaceutical Sciences, University of Bologna, Via Belmeloro 6, 40126 Bologna, Italy
Received 30 August 2007; revised 27 September 2007; accepted 29 September 2007
Available online 4 October 2007
Abstract—The complex etiology of Alzheimer’s disease (AD) prompts scientists to develop multifunctional compounds to combat
causes and symptoms of such neurodegeneration. To this aim we designed, synthesized, and tested a series of compounds by intro-
ducing halophenylalkylamidic functions on the scaffold of AP2238, which is a dual binding site acetylcholinesterase inhibitor. The
inhibitory activity was successfully extended to the beta-site amyloid precursor protein cleavage enzyme, leading to the discovery of
a potent inhibitor of this enzyme (3) and affording multifunctional compounds (2, 6, 8) for the treatment of AD.
Ó 2007 Published by Elsevier Ltd.
Alzheimer’s disease (AD), a progressive, degenerative
disorder of the brain, is believed to be the most common
cause of dementia among the elderly. AD is associated
with a loss of the presynaptic markers of the cholinergic
system in the brain areas related to memory and learn-
ing,1 and is characterized by the presence of amyloid
deposits and neurofibrillary tangles in the brain of af-
flicted individuals. The disease appears to have a heter-
ogeneous etiology, but the deposition of beta-amyloid
(Ab) peptides (amyloid plaques) in the brain is hypoth-
esized to be the root cause of neuronal cell death in AD
patients and to play a key role in the progression of the
disease.2,3
On therapeutic fronts, BACE1 inhibitors need to be less
peptidic or nonpeptidic in character to achieve good oral
bioavailability and, more important, CNS penetration.
Up to now most of the drugs approved for AD treat-
ment are AChE inhibitors, which are able to enhance
cholinergic neurotransmission by increasing acetylcho-
line (ACh) availability in the synaptic cleft. Considering
the mechanism of action of these drugs, they are not
expected to interfere with the neurodegenerative cascade
of the disease, but only to temporarily mitigate some of
the symptoms. However, some studies have shown that
they can be effective over a longer period.8 Moreover,
AChE plays an important role in Ab deposition: besides
its catalytic function, AChE also acts as a promoter of
Ab fibril formation, this effect being independent from
its normal hydrolyzing activity and associated with the
peripheral anionic site (PAS) of the enzyme.9
A very attractive approach to lowering Ab is to inhibit
b-secretase, generally referred to as BACE1 (beta-site
amyloid precursor protein [APP] cleavage enzyme), a
transmembrane aspartyl protease responsible for N-ter-
minal cleavage of the APP leading to the production of
the Ab peptide.4 Thus, BACE1 has become a target5 of
significant interest for industrial and academic investiga-
tion and there is important research underway in the
development of new inhibitors for this enzyme.6,7
Due to the complexity of AD and the involvement of
different enzymes in its progression, the modulation of
a single protein might not be sufficient to produce the
desired efficacy. In the light of this, researchers are
now turning to the design of compounds able to simul-
taneously hit different targets.10,11 In a previous work12
we designed and synthesized a dual binding site AChE
inhibitor, AP2238 (Fig. 1), which is able to simulta-
neously contact both the central and the peripheral
anionic sites and which showed an IC50 of 44.5 nM
Keywords: Alzheimer; Multi-target; hAChE; BACE1.
*
Corresponding author. Tel.: +39 0 512099722; fax: +39 0 512099734;
e-mail: lorna.piazzi@unibo.it
against hAChE activity. From a further SAR study13
a
0960-894X/$ - see front matter Ó 2007 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2007.09.100

424
L. Piazzi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 423–426
and 8, 10-fold less potent than 1, still maintained a fairly
good activity, comparable to the monomethoxycouma-
rins reported in a previous paper.13 The considerable
O
R
N
decrease in activity shown for compound
5
(IC50 = 267 lM), with a trans-3,4-dichlorocinnamic sub-
stituent, could indicate that molecules carrying bulky
groups are not allowed to penetrate into the hAChE
gorge to establish a proper interaction, because the sub-
stituent in that position might detrimentally interact
with hAChE PAS residues. Regarding the activity to-
ward BACE1, the data reported in Table 1 show that
all compounds were potent inhibitors, but activity
seemed not to be influenced either by the position of
the substituent on the coumarin nucleus, or by the differ-
ent halogen on the phenyl ring. This portion of the mol-
ecule was crucial for efficient BACE1 inhibition, as
confirmed by the lower potency of the reference com-
pound 1, which lacked this group and showed inhibition
of 42.33% at 210 nM. Compound 3 was the most active
of the series, with an IC50 of 99 nM.
R = Me AP2238
R = Et 1
Figure 1.
new compound emerged, 1 (AP2243, Fig. 1), with an
ethyl instead of a methyl group on the basic nitrogen,
showing an improved IC50 (18.3 nM).
In this letter, we report new amidic nonpeptidic deriva-
tives in which the structure of 1 (for AChE activity) was
maintained, whereas the methoxy groups of the couma-
rin moiety were alternately substituted by an amidic
chain to extend the activity to BACE1. Therefore, we
introduced halophenylalkylamidic functions in positions
6 or 7 of the coumarin moiety, choosing a dihalophenyl
acid because this moiety emerged as a leitmotif in differ-
While the binding mode of these derivatives at hAChE
gorge resembles that of the parent compound,12 docking
outcomes concerning an inhibitor-BACE1 complex are
here reported. Docking simulations were carried out
using GOLD18 and Dock19 suites and the outcomes
grouped together and clusterized with AClAP20,21
according to the so-called ‘holistic approach’.21 Docking
calculations revealed that among all the investigated
compounds (see Supplementary data) a similar trend in
the binding mode was observed. In the light of this, the
binding mode of the most potent BACE1 inhibitor, 3,
is here discussed. The flexibility of the substituent in po-
sition 6 on the coumarin moiety led to heterogeneous
docking poses. Nevertheless, the availability of the clus-
tering method was very helpful during the analysis of
docking solutions. Eight hundred docking poses for
compound 3 were obtained as described in Supplemen-
tary data. The total number of clusters obtained was 74
but, according to the Chauvenet criterion implemented
in AClAP,20,21 only two clusters were significantly popu-
lated. The ascertainment of the ligand–protein com-
plexes, supported also by analysis of the docking
solution for the other molecules of the series, led us to
consider the binding mode proposed in Figure 1 as plau-
sible. Here, the representative pose of the most populated
cluster is shown after the Dock6.1 Amber rescoring.18 It
can be seen that the protonated nitrogen of the ligand is
able to reach the acid environment formed by the cata-
lytic diad Asp32 and Asp228; its position inside the ac-
tive site resembles that of protonated nitrogens of
classical hydroxyethylamine inhibitors. The N-benzyl-
ethyl moiety seems particularly suitable for simulta-
neously fitting the S1 and S10 pockets. In Figure 2, the
N-benzyl group interacts with Tyr198 and Ile226,
whereas the N-ethyl portion lies on the aromatic ring
of Tyr71. In the present docking study, this interaction
pattern has also been observed in an inverted fashion
having the N-benzyl substituent T-shape interacting with
Tyr71 and Phe118 and the N-ethyl moiety lying on the
S10 pocket. The coumarin core with its 3-phenyl substitu-
ent is embedded into the S2 pocket of the enzyme, with
the carbonylic group being able to H-bond the side chain
ent BACE1 inhibitors reported in the literature.14,15
A
new series of potential multi-target compounds (2–9)
for AD was synthesized, whose structures are reported
in Table 1.
According to Scheme 1, compounds 2–5 were synthe-
sized starting from 3-{4-[(benzylethylamino) methyl]-
phenyl}-6-methoxychromen-2-one, and compounds 6–
9 from 3-{4-[(benzylethylamino)methyl]phenyl}-7-meth-
oxychromen-2-one, which were treated with 48% HBr to
afford the hydroxy derivatives 10 and 11, respectively.
The subsequent substitution with 2-Boc-aminoethyl bro-
mide gave the intermediates 12 and 13, respectively. Re-
moval of the protecting group (Boc) by means of
CF3COOH released the primary amino functions (14
and 15). The final amidic derivatives (2–9) were formed
by reaction with DCC activated (3,5-difluorophenyl)ace-
tic acid, 3,5-difluorocinnamic acid, 3,4-difluorohydro-
cinnamic acid or 3,4-dichlorocinnamic acid, via a
parallel synthesis procedure: in two series of four dis-
tinct reactors, (3,5-difluorophenyl)acetic acid, 3,5-diflu-
orocinnamic acid, 3,4-difluorohydrocinnamic acid, and
3,4-dichlorocinnamic acid (1.3 equiv) were dissolved in
CH2Cl2 under N2 atmosphere. Then, DCC (1.4 equiv)
was added to each reactor. Amino derivatives 14 (in
the first series) or 15 (in the second series) (1.0 equiv)
were added at 0 °C, and each mixture was stirred at
room temperature for 2 h under N2. The DCU was fil-
tered off from each solution. Purification of each crude
product by flash chromatography yielded the corre-
sponding amides 2–9 (Tabel 1).
The inhibitory activities of the newly synthesized com-
pounds were studied against hAChE, using the method
of Ellman16 to determine the rate of hydrolysis of acet-
ylthiocholine, and against BACE1, using a spectrofluo-
rometric method.17
The results reported in Table 1 for all tested compounds
showed a decrease in activity toward hAChE with re-
spect to the reference compound 1. Compounds 2, 6,

L. Piazzi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 423–426
425
Table 1. Structures of studied compounds and their hAChE and BACE1 inhibitory activities, expressed as IC50
O
O
7
6
O
R
O
NH
N
Compound
Position
—
R
IC50 SEMa (lM)
BACE1
hAChE
1
—
0.018 0.003
ꢀ0.238b
F
2
3
6
6
0.551 0.049
0.149 0.002
F
F
7.16 0.87
0.099 0.005
F
F
F
4
5
6
6
2.77 0.25
267 103
0.116 0.001
0.133 0.004
Cl
Cl
F
6
7
0.181 0.010
0.150 0.007
F
F
7
8
9
7
7
7
4.57 0.39
0.327 0.034
4.23 0.49
0.121 0.002
0.114 0.007
0.151 0.001
F
F
F
Cl
Cl
a Human recombinant AChE and BACE1 were used. IC50 values represent the concentration of inhibitor required to decrease enzyme activity by
50% and are means of two independent measurements, each performed in triplicate (SEM, standard error of the mean).
b 42.33% inhibition at 201.66 nM.
of Arg235. The presence of a substituent in positions 6 or
O
O
O
O
7 on the coumarin nucleus seems to equally contribute to
the relative binding affinity of this series of molecules.
This might be due to both the flexibility of the bridging
aminoethoxy segment and to the relative structural per-
mittivity of the enzyme portion. The orientation of the
6-substituent might be stabilized by the interaction of
the amidic nitrogen with the backbone of the 1OS loop,
Gly11. Moreover, the 3,5-difluorophenyl portion might
be harbored into a peripheral cationic spot formed by
residues Lys9, Arg307, Lys321, Pro160, and Val309.
O
H
O
a
b
N
Ph
N
Ph
10-11
O
O
O
O
O
O
c
H2N
14-15
HN
d
N
Ph
12-13
N
Ph
BOC
O
O
O
R
O
HN
N
Ph
2-9
In conclusion, the introduction of an halophenylalky-
lamidic function on the scaffold of 1 allowed us to obtain
potent BACE1 inhibitors. In particular, compound 3 was
the best of the series, showing an IC50 of 99 nM. Due to
Scheme 1. aSynthesis of the studied compounds. aReagents: (a) HBr;
(b) 2-(Boc-amino)ethyl bromide; (c) CF3COOH; (d) DCC, (3,5-
difluorophenyl)acetic acid, 3,5-difluorocinnamic acid, 3,4-difluorodi-
hydrocinnamic acid or 3,4-dichlorocinnamic acid.

426
L. Piazzi et al. / Bioorg. Med. Chem. Lett. 18 (2008) 423–426
article can be found, in the online version, at doi:-
doi:10.1016/j.bmcl.2007.09.100.
References and notes
1. Terry, A. V., Jr.; Buccafusco, J. J. J. Pharmacol. Exp.
Ther. 2003, 306, 821.
2. Selkoe, D. J. Annu. Rev. Neurosci. 1994, 17, 489.
3. Hardy, J.; Selkoe, D. J. Science 2002, 297, 353.
4. Selkoe, D. J. Nature (London) 1999, 399A, 23.
5. Citron, M. J. Neurosci. Res. 2002, 70, 373.
6. Hussain, I. IDrugs 2004, 7, 653.
7. Hills, I. D.; Vacca, J. P. Curr. Opin. Drug Discov. Devel.
2007, 10, 383.
8. Bullock, R.; Dengiz, A. Int. J. Clin. Pract. 2005, 59, 817.
9. De Ferrari, G. V.; Canales, M. A.; Shin, I.; Weiner, L.
M.; Silman, I.; Inestrosa, N. C. Biochemistry 2001, 40,
10447.
Figure 2. Proposed binding mode for compound 3 (carbon atoms in
rose) docked at the active site (cyan) of BACE1. S1 and S10 enzyme
subsites are sketched by dotted red lines. Amino acids are indicated as
single-letter code.
10. Morphy, R.; Rankovic, Z. J. Med. Chem. 2005, 48, 6523.
11. Morphy, R.; Rankovic, Z. Fragment, network biology
and designing multiple ligands. Drug Discov. Today, 2007,
12, 156.
12. Piazzi, L.; Rampa, A.; Bisi, A.; Gobbi, S.; Belluti, F.;
Cavalli, A.; Bartolini, M.; Andrisano, V.; Valenti, P.;
Recanatini, M. J. Med. Chem. 2003, 46, 2279.
13. Piazzi, L.; Cavalli, A.; Belluti, F.; Bisi, A.; Gobbi, S.;
Rizzo, S.; Bartolini, M.; Andrisano, V.; Recanatini, M.;
Rampa, A. J. Med. Chem. 2007, 50, 4250.
the multifactorial nature of AD, molecules that modu-
late the activity of a single protein target are unable to
significantly modify the progression of the disease. Fol-
lowing this new paradigm, here we have reported the first
dual AChE/BACE1 inhibitors. Promising hits proved to
be compounds 2, 6, and 8, which turned out to inhibit
both enzymes with sub-micromolar affinity, as required
for a good multi-target-directed drug candidate.
14. Drug Data Report, 2002, 24, 119, 309, 597, 885.
15. Guo, T.; Hobbs, D. W. Curr. Med. Chem. 2006, 13, 1811.
16. Ellman, G. L.; Courtney, K. D.; Andres, V.; Featherstone,
R. M. Biochem. Pharmacol. 1961, 7, 88.
Acknowledgments
17. Mancini, F.; Naldi, M.; Cavrini, V.; Andrisano, V. Anal.
Bioanal.Chem. 2007, 388, 1175.
18. Jones, G.; Willet, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
J. Mol. Biol. 1997, 267, 727.
This work was supported by MUR, Rome (Grant No.
FIRB RBNE03FH5Y). We thank Ms. Claudia Ferroni
for her technical assistance.
19. Moustakas, D. T.; Lang, P. T.; Pegg, S.; Pettersen, E.;
Kuntz, I. D.; Brooijmans, N.; Rizzo, R. C. J. Comput.
Aided Mol. Des. 2006, 20, 601.
Supplementary data
20. Bottegoni, G.; Cavalli, A.; Recanatini, M. J. Chem. Inf.
Model. 2006, 46, 852.
21. Bottegoni, G.; Rocchia, W.; Recanatini, M.; Cavalli, A.
AClAP, Bioinformatics 2006, 22, e58.
Full experimental procedures for the synthesis, compu-
tational studies, and biological evaluation of the studied
compounds. Supplementary data associated with this