Versatility of the Curcumin Scaffold: Discovery of Potent and Balanced Dual BACE-1 and GSK-3β Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.5b00894

The multitarget approach has gained increasing acceptance as a useful tool to address complex and multifactorial maladies such as Alzheime?s disease (AD). The concurrent inhibition of the validated AD targets β-secretase (BACE-1) and glycogen synthase kin

Full text of DOI:10.1021/acs.jmedchem.5b00894

Subscriber access provided by CMU Libraries - http://library.cmich.edu
Article
The Versatility of the Curcumin Scaffold: Discovery of
Potent and Balanced Dual BACE-1 and GSK-3# Inhibitors
Rita M.C. Di Martino, Angela De Simone, Vincenza Andrisano, Paola Bisignano,
Alessandra Bisi, Silvia Gobbi, Angela Rampa, Romana Fato, Christian Bergamini,
Daniel I. Perez, Ana Martínez, Giovanni Bottegoni, Andrea Cavalli, and Federica Belluti
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00894 • Publication Date (Web): 22 Dec 2015
Downloaded from http://pubs.acs.org on December 23, 2015
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 41
Journal of Medicinal Chemistry
1
2
3
4
The Versatility of the Curcumin Scaffold: Discovery of Potent and Balanced Dual BACE-1
5
6
and GSK-3β Inhibitors
7
8
9
Rita Maria Concetta Di Martino,1 Angela De Simone,2 Vincenza Andrisano,2 Paola Bisignano,3
Alessandra Bisi,1 Silvia Gobbi,1 Angela Rampa,1 Romana Fato,1 Christian Bergamini,1 Daniel I.
Perez,4 Ana Martinez,4 Giovanni Bottegoni,3 Andrea Cavalli,1,3 Federica Belluti1*
10
11
12
13
14
15
16
17
18
1Department of Pharmacy and Biotechnology, Alma Mater Studiorum - University of Bologna, Via
Belmeloro 6, 40126 Bologna (Italy);
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
2Department for Life Quality Studies, Alma Mater Studiorum - University of Bologna, Corso
D’Augusto 237, 47921 Rimini (Italy);
3Istituto Italiano di Tecnologia, D3, via Morego 30, 16163 Genova (Italy);
4Centro de Investigaciones Biologicas, CSIC, Ramiro de Maetzu 9, 28040 Madrid (Spain).
Abstract
Multi-target approach has gained increasing acceptance as useful tool to address complex and
multifactorial maladies such as Alzheimer’s disease (AD). The concurrent inhibition of the
validated AD targets β-secretase (BACE-1) and glycogen synthase kinase-3β (GSK-3β), by
attacking both beta-amyloid and tau protein cascades, has been identified as a promising AD
therapeutic strategy. In our study, curcumin was identified as lead compound for the simultaneous
inhibition of both targets; therefore, synthetic efforts were dedicated to obtaining a small library of
novel curcumin-based analogues, and a number of potent and balanced dual-target inhibitors were
obtained. In particular, 2, 6 and 7 emerged as promising drug candidates endowed with
neuroprotective potential and brain permeability. Notably, for some new compounds the
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 41
1
2
3
4
5
6
symmetrical diketo and the β-keto-enol tautomeric forms were purposely isolated and tested in
vitro, allowing us to gain insight into the key requirements for BACE-1 and GSK-3β inhibition.
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
Introduction
Alzheimer’s disease (AD), the most common form of dementia in industrialized nations among
the elderly, is a progressive degenerative disorder, characterized by massive neuronal death and
synaptic degeneration, that ultimately leads to the collapse of the neural networks essential for
cognition and memory. It is the 6th leading cause of death, currently affecting more than 44 million
people worldwide and, due to its debilitating nature, causes an enormous financial and emotional
stress on patients and caregivers. Considering that, worldwide, the population is rapidly aging, the
number of AD patients is projected to reach 116 million by 2050.1 The current FDA-approved
therapies for moderate to severe AD provide only temporary and incomplete symptomatic relief and
only represent a palliative tool to slow down the clinical course of the disease.2 The lack of truly
effective drugs is related to the multifactorial nature of the disease, in which several impaired
molecular pathways interact in a feed-forward loop.3,4
An aberrant protein processing is a distinctive feature of AD, in which amyloid β (Aβ) peptide
and abnormally hyperphosphorylated tau protein, upon misfolding, self-assembly and accumulate in
the brain as aggregates, namely amyloid senile plaques (SP) and neurofibrillary tangles (NFT),
respectively.5 These assemblies represent the most relevant histopathological hallmarks of the
disease and have been considered to play crucial roles in its pathogenesis. Through different
molecular mechanisms these aggregates trigger a cascade of biological processes, among others,
amyloid and tau cascades, ultimately culminating in neuronal cell death, brain atrophy, and
cognitive decline.
Research programs focused on the amyloid cascade still occupy a relevant place, as documented
by the substantial number of compounds entered in clinical trials,6 together with the large amount of
2
ACS Paragon Plus Environment

Page 3 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
published manuscripts. Aβ peptide is generated through a sequential proteolysis of the amyloid β
protein precursor (APP) catalyzed by β- and γ-secretases.7 In this context, the inhibition of β-
secretase, an aspartyl protease also known as β-site APP cleaving enzyme (or BACE-1), regulating
the first and rate-limiting step of the APP processing, could efficiently block the production of Aβ
and of other post-translational products.8 The higher BACE-1 expression and activity detected in
AD patients’ brain are consistent with the elevated Aβ levels.9 The availability of BACE-1 crystal
structure greatly enabled the drug development process in this field. This aspartyl protease is
characterized by a bilobal structure, the active site is a long cleft for substrate recognition with a
flap region that, descending over the top, pins the substrate nearby two catalytic aspartic residues at
the site of bond hydrolysis.10 Interestingly, a molecular interplay between Aβ and tau in causing
synergic toxicity is a recent and intriguing finding, even if the molecular basis of the process still
remain to be fully deciphered.11
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
Under normal physiological conditions tau is associated with microtubules and, preventing their
dynamic instability, contributes to morphogenesis of neurons.12 The strong correlation between tau
hyperphosphorylation and AD pathology focuses the attention on tau kinases and, glycogen
synthase kinase-3β (GSK-3β), a multi-tasking serine / threonine kinase largely expressed in central
nervous system (CNS), proved to play a significant role in regulating tau phosphorylation under
both physiological and pathological conditions.13 In particular, GSK-3β dysregulation is believed to
contribute, through a large number of cellular processes, to the etiology of chronic conditions such
as diabetes, cancer, and CNS disorders i.e. schizophrenia and AD.14 GSK-3β induces tau
phosphorylation mainly at Ser199, Ser396, Ser413, causes tau disjunction from the microtubules
and aggregation as NFT.15 Over the last decade, the increased interest in GSK-3β led to the
discovery of a number of inhibitors, based on chemically different molecular scaffolds, and acting
with different mechanisms of action, such as ATP competition, allosteric modulation, and enzyme
irreversible inhibition.16 The last two approaches allowed to minimize the undesirable off-target
effects. In particular, an irreversible inhibition, by selective targeting of a non-conserved cysteine
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 41
1
2
3
4
5
6
7
8
9
residue (Cys199) in the ATP-binding site, was reported as a promising strategy for obtaining useful
pharmacological tools.17,18 Interestingly, the cross-talk between GSK-3β and Aβ suggested GSK-3β
as molecular link between Aβ and tau. The GSK-3β pathological activation by Aβ, through the
prevention of the inhibitory phosphorylation of this kinase, leads to an increment of tau
phosphorylation.19,20 Furthermore, inhibition of this kinase decreases Aβ production and Aβ-
induced neurotoxicity by reducing BACE-1 cleavage of APP.14
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
Disease-modifying therapeutics, able to hit AD underlying physiopathology, would represent an
ideal tool to control both onset and progression of the neurodegenerative process.21,22 Recently, it
has become evident that the concurrent modulation of several targets involved in this complex
disease could represent an ideal strategy for achieving an effective treatment.23,24 The idea is to
devise multi-target agents that might provide a better efficacy profile, compared with a single-target
therapeutic.25 In this respect, a crucial issue of the multi-target drug discovery is the achievement, in
a single chemical entity, of balanced potency toward the selected targets .26 In this scenario, a multi-
target approach, based on the simultaneous inhibition of disease-modifying AD relevant targets
BACE-1 and GSK-3β, offers promises to achieve a successful AD treatment.27 To further
corroborate this challenging and still underexplored drug discovery hypothesis, and as continuation
of our research studies aimed at the discovery of natural products-inspired compounds as
multipotent AD drug candidates,28,29 we rationally designed a new series of dual BACE-1 and GSK-
3β inhibitors based on the curcumin scaffold.
Natural products proved to be an excellent source of lead compounds for drug discovery.30 These
small molecules, synthesized by plant kingdom, are considered as “privileged structures” because
they have evolved in the natural selection process to achieve optimal interactions with a number of
biological targets.31 The polyphenol curcumin (1a, Figure 1), the primary bioactive compound
found in the rhizomes of Curcuma longa L.,32 is one of the most thoroughly investigated natural
product, as documented by studies claiming its efficacy and safety for both prevention and
treatment of various disorders.33,34 This pharmacological pleiotropic behaviour has been proposed to
4
ACS Paragon Plus Environment

Page 5 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
be strictly related to its mechanism of action that consists in a synergistic binding to different target
proteins involved in related signalling pathways. For instance, molecular basis of curcumin-
mediated neuroprotection revealed the participation of a wide range of AD interrelated pathways,
namely aggregation of Aβ and tau proteins, oxidative stress, and neuroinflammation.35 Moreover,
from a medicinal chemistry standpoint, the curcumin pharmacophore proved to be an excellent lead
for the design of analogues with improved biological profile.36
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
Several investigations have been focused on curcumin tautomeric behavior, from a theoretical
point of view, and the structure of curcumin in solution has been represented as an interconverting
mixture of the symmetric diketo tautomer 1a (1E,6E-1,7-bis-4-hydroxy-3-methoxyphenylhepta-1,6-
diene-3,5-dione) and the asymmetric β-keto-enol tautomer 1b (1E,4Z,6E-5-hydroxy-1,7-bis-4-
hydroxy-3-methoxyphenylhepta-1,4,6-trien-3-one), (Figure 1). More recent theoretical and
spectroscopic studies clearly provided evidences that 1b is the predominant form in a wide range of
organic solvents and buffered solutions due to the stabilizing effect of the intramolecular H-bond on
the ,-unsaturated-keto-enol moiety (Figure 1), property that may play a critical role in
determining its affinity to biological targets.37-38
Rational design of dual BACE-1 and GSK-3β inhibitors
Notwithstanding the fact that BACE-1 and GSK-3β are structurally unrelated enzymes and
greatly differ in both three-dimensional structures and binding site topologies, we rationally
envisaged the 5-hydroxy-1,7-diarylhepta-1,4,6-trien-3-one framework (main backbone of 1b), as
appropriate structural element for their simultaneous modulation. In particular, the β-keto-enol
central core of curcumin looked suitable for the interaction with BACE-1 catalytic dyad (Asp32 and
Asp228), whereas the highly electrophilic α,β-unsaturated carbonyl system, a well-known Michael
acceptor system,39 could covalently interact with the crucial Cys199 residue of GSK-3β.
To substantiate this hypothesis, the interactions of curcumin 1b with the binding pocket of both
targets were investigated by means of docking simulations (Figures 2a and 2b). Figure 2a showed
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 41
1
2
3
4
5
6
7
8
9
how, in BACE-1 catalytic pocket, the central core was hydrogen bonded to the crucial Asp32 and
Asp228 residues, while the side aryl rings established contacts with Tyr198, Lys224, and Tyr71.
Figure 2b reported the predicted binding affinities of 1b at GSK-3β binding pocket, that were in
good agreement with a previously proposed binding mode.40 In particular, a number of significant
interactions were established, and, namely, the central β-keto-enol function participated in a H-
bond with Tyr134 and Val135, and the two aryl functions contacted Lys85, Glu97, and Arg141,
Glu173, respectively.
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
Furthermore, we tested the naturally occurring curcumin (1b) and its simplified analogue bis-
demethoxycurcumin (4) for their ability to inhibit both BACE-1 and GSK-3β enzymes (Table 1).
With regards to BACE-1 inhibition, curcumin proved to have in moderate potency, while low-
micromolar activity was observed for 4, confirming previously reported data.41 In the GSK-3β
inhibition assay, both compounds proved to analogously modulate the enzyme (micromolar IC50
values). In summary, 4 showed a promising dual BACE-1 / GSK-3β inhibitory profile (2.54 µM and
8.39 µM on BACE-1 and GSK-3β, respectively). Taken together, these initial investigations
confirmed the suitability of the curcumin scaffold for the development of effective dual BACE-1
and GSK-3β inhibitors.
To explore the chemical space of the two targets and perform a preliminary structure-activity
relationship (SAR) study, a small library of curcumin-based analogues (Figure 1) was designed and
synthesized. Different moieties were introduced on the side aryl functions of the main scaffold,
while the central heptadienone fragment was retained. The choice of the substituents was mainly
addressed to favor the crossing of the blood-brain barrier (BBB), as this pharmacokinetic property
represents an essential requirement for drugs targeting CNS. The new curcumin-based analogues
were primarily screened on BACE-1 and GSK-3β enzymes to determine their inhibitory profiles
(Tables 1 and 2). Subsequently, for the most promising compounds, in terms of inhibitory activities
(high potency against one target or balanced inhibitory profile on both targets), the AD-modifying
profile, namely neuroprotection, potential neurotoxicity, and ability to cross the BBB, were also
6
ACS Paragon Plus Environment

Page 7 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
investigated. Moreover, docking simulations were performed, in order to study the general
positioning of this new series within the active site of both targets and to assess specific interactions
with the residues of the catalytic pockets. To this aim, we selected 8 and 5 as the most potent
BACE-1 and GSK-3β inhibitors of the series, respectively, and 2, 4, 6, and 7 as low-micromolar
dualistic inhibitors.
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
Chemistry
The synthetic route for obtaining the target curcumin-based compounds (2-8, 11, 12) is outlined
in Schemes 1 and 2, applying the Pabon reaction.42 In summary, pentane-2,4-dione was complexed
with B2O3 in ethyl acetate (EtOAc) to avoid the methylene-centred reactivity toward the
Knoevenagel reaction and to favour the nucleophilic attack at the side methyl groups. The boric
complex was then condensed with the suitable aldehyde and then a step-wise addition of n-
tributylborate and n-butylamine was carried out. Acidic treatment allowed the complex to
dissociate. This reaction has been reported to give curcumin as β-keto-enol tautomer. The synthesis
of the asymmetric curcumin-based analogues (2, 3, 6, 7) could be performed by condensing two
different aldehydes, obtaining thus a complex mixture of compounds, among which the
asymmetrical and the two symmetrical curcuminoids, including the semi-reaction products (mono-
aryl curcumin) products. Consequently, obtaining the desired compounds in a good yield and purity
grade would require several purification processes (column chromatography). To avoid this
drawback, we performed the synthesis of the asymmetric curcumin analogues (Scheme 1) via a
two-steps strategy, in which the mono-aryl curcumin synthetic intermediates were first prepared
under the classical reaction condition, employing vanillin or 4-benzyloxybenzaldehyde (17 and 18,
respectively). Subsequently, a second Pabon reaction, with a suitable aldehyde, allowed to obtain
the desired final compounds, maintaining the β-keto-enol tautomeric form. Scheme 2 reports the
synthetic one-pot Pabon reaction procedure for obtaining the symmetrical analogues, by using a
stoichiometric ratio of 1:1.8 for acetylacetone and the selected substituted benzaldehyde (4, 5, 8, 11,
12).
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 41
1
2
3
4
5
6
7
8
9
The Williamson ether synthesis (Scheme 3) between a phenol-key intermediate (4) and a selected
benzyl halide, in the presence of K2CO3 as base, gave the desired products in a mixtures of β-keto-
enol (8-10) and symmetric diketo tautomers (14-16). In particular, the alkaline reaction conditions,
induced a transformation of the β-keto-enol tautomer into the corresponding diketo one. Moreover,
the employment of a sterically demanding alkylating reagent, was essential for obtaining both the
tautomeric ethers into the reaction mixture. Additionally, these encumbering chemical features
provided each reaction product with a fairly different chromatographic behaviour; and, as a
consequence, each single tautomers could then be successfully and carefully isolated from the crude
reaction by flash column chromatography, and further purified by crystallization from a suitable
solvent, thus obtaining the target compound in high purity grade. Interestingly, when methyliodide
was used as alkylating reagent (Scheme 3), it was not possible to obtain the corresponding
tautomers 5 and 13 as chromatographically different compounds and, to avoid separation
difficulties, reaction time was lengthened in order to convert 5 into 13; a by-product was also
separated ad identified. This type of reaction also allowed to study the stability of the synthesized
derivatives (see Chemical Stability Studies).
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
Results and Discussion
BACE-1 and GSK-3β Inhibition
BACE-1 inhibition
The ability to inhibit BACE-1 enzyme activity was investigated by means of a biochemical assay
performed using the fluorescence resonance energy transfer (FRET) methodology43 and using
curcumin as reference compound. The results are reported in Tables 1 and 2. All the tested
compounds 2-16 turned out to effectively inhibit BACE-1, more than 1b (in our in vitro test, it does
not inhibit this enzyme up to 3 µM) with IC50 ranging from nanomolar to low micromolar values.
These good potencies confirmed the suitability of the curcumin main framework for BACE-1
inhibition, substantiating the design rationale for this target, even if the nature of the substituents on
8
ACS Paragon Plus Environment

Page 9 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the side aryl functions seemed to have an intense effect on activity. The structurally closely related
2 and 3, bearing curcumin’s 4-hydroxy-3-methoxyphenyl as A ring, and a 4-benzyloxyphenyl or
para-tolyl as B ring, respectively, proved to effectively inhibit BACE-1 (IC50 values of 0.97 and
0.14 µM, respectively). The corresponding bis-para-benzyloxyphenyl symmetrical analogue 8
showed high potency, and, with an IC50 value of 40 nM, was the most active compound of the
series. A remarkable drop in potency (two orders of magnitude) was observed for the corresponding
diketo tautomer (14). Docked pose of 8 at BACE-1 binding site (Figure S1a in Supporting
Information) was in good agreement with the reported activity, as the molecule spanned the enzyme
active site and the benzyloxy rings pointed toward both the S and the S’ subpockets, binding the
hydrophobic region described by the side chains of aromatic residues (Tyr198) and the flap region.
From this structure-based docking simulation, it was evident that, to maintain this optimal binding
mode, the benzyl rings of 8-based analogues could only be decorated with small substituents, such
as a fluorine atom. The latter, when strategically positioned, has been reported to positively affect
the biological profile in terms of enhanced binding interaction, metabolic stability, and selective
reactivity44 allowing the development of effective BACE-1 inhibitors.45,28 Symmetrically
fluorinated 8-analogues were synthetized and a lower potency (one order of magnitude) relative to 8
was observed for compounds 9 and 10 (2,4-di-F and 3,5-di-substituted analogues, IC50 = 0.40 and
0.39 µM, respectively), and a further drop in activity was documented for 11 and 12 (4-F and 3-F
substituted analogues, IC50 = 1.04 and 1.08 µM, respectively). For this subset, the corresponding
diketo tautomers of the di-fluorinated 9 and 10 (15 and 16), as the most active compounds of the
series, showed only a moderate decrement of activity. When one or both benzyloxy moieties of 8
were replaced by methoxy and hydroxy (4-7), low micromolar inhibitory potencies were observed.
Analogously to 14, the 5-diketo tautomer (13) was less active than the corresponding β-keto-enol.
Taken together, these data pointed out that the structural modifications on 8 had, in general,
detrimental effects on BACE-1 inhibiting activity.
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
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 41
1
2
3
4
GSK-3β Inhibition
5
6
7
8
A luminescent assay was employed for the evaluation of the GSK-3β inhibitory potency.46 The
curcumin-based analogues (2-16) showed a narrower GSK-3β inhibition range compared to that
observed for BACE-1 (IC50 values from 0.53 to 16.99 µM). All the tested compounds showed
micromolar activities, except for the para-methoxy substituted 5, with an IC50 value of 0.53 µM,
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
that proved to be the most active in the series, more potent than 1b (IC50 = 17.95 µM) and 4 (IC50 =
8.39 µM). Good GSK-3β inhibition was also obtained with the replacement of a side aryl ring of
curcumin with a para-benzyloxyphenyl or a para-tolyl functions (2 and 3, IC50 = 0.90 and 2.09 µM,
respectively). Moreover, 8, 6, and 7, all bearing a para-benzyloxyphenyl moiety, showed similar
inhibitory potencies (around 2 µM). A lowered inhibitory potency was observed for the 8-based
fluorinated subset (IC50 values ranging from 8.30 to 16.99 µM) that proved to be slightly more
active or as active as 1b. Concerning the diketo tautomers, only for 13 a notable divergence in
activity was observed with respect to the β-keto-enol counterpart, for the other compounds of this
series bearing para-benzyloxyphenyl groups (14, 15 and 16) only a moderate decrease in activity
was observed.
In the light of these results, among the tested compounds 8 and 5 proved to be the most potent
BACE-1 and GSK-3β inhibitors, respectively. A well-balanced dual low-micromolar inhibition
profile was observed for 2, 6 and 7, thus offering promises for gaining ground in the development
of curcumin-based multitarget drug candidates for AD treatment.
Computer-assisted studies
Aimed at gaining insight into the molecular determinants of the curcumin-based analogues
involved in effective interactions with BACE-1 and GSK-3β binding sites, docking studies were
performed. To this aim, 8 and 5 were selected as highly active principally on one of the two targets,
and 6 and 7 as balanced inhibitors on both targets (see Supporting Information for the binding mode
description and for Figures S1a-d and S2a-d displaying docking predictions). We subsequently
10
ACS Paragon Plus Environment

Page 11 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carried out molecular docking simulations to study the interactions of the balanced low-micromolar
inhibitor 2 with the catalytic pocket of both targets. (Figures S3a and S3b in the Supporting
Information)
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
Covalent docking simulation on GSK-3β
Taking into account the frequently claimed strong correlation between compound’s reactivity and
bioactivity, and the highly and well established electrophilic properties of the curcumin
pharmacophore, a covalent docking was accomplished in which a thia-Michael reaction occurred by
nucleophilic attack of GSK-3β Cys199 residue on the reactive α,β-unsaturated carbonyl function of
5 (Figure 3). It can be speculated that specific non-covalent interactions of the molecule with the
binding pocket of the target are critical for the optimal orientation of the electrophile
pharmacophore toward a specific protein nucleophile residue, thus increasing the speed and
selectivity of the interaction (covalent bond).47
Thiol trapping assay
To support the covalent docking study, a spectroscopic 1H-NMR-based thiol trapping assay was
performed. This procedure has been described for several electrophilic compounds, among them
curcumin,48 that was reported to irreversibly and quickly react with cysteamine in DMSO-d6 to
afford the corresponding bis-1,7-thia-Michael adduct, as documented by the disappearance of the
olefin signals on the 1H-NMR spectra.49 Compound 5 underwent a thiol-trapping reaction by
employing a mixture of 5 / cysteamine in a 1:2 stoichiometric ratio,48 affording the corresponding
thia-Michael adduct in very short time as documented in Figure 4. Analogously, competitive
experiments were also carried out on 2, 8 and 6; longer times of reaction were observed, with
respect to 5, with the following reactivity order: 5 > 2 > 6 ~ 8, suggesting a potential detrimental
effect on reactivity elicited by the benzyloxy function, partially compensated by the presence of the
4-OH, 3-OCH3-aryl ring (2).
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 41
1
2
3
4
Chemical Stability Studies
5
6
7
8
A number of studies provided evidences about curcumin’s poor stability in aqueous solutions at
physiological pH and several degradation products were identified.50, 51 The chemical stability of 5
as β-keto-enol form and its diketo tautomeric counterpart 13, together with their potential
interconversion, were studied by RP-HPLC. For 5, in the working conditions, no degradation or
tautomeric conversion was observed. On the contrary, by analyzing the DMSO solution of 13 the
formation of a degradation product was noted. This derivative was separated from the mixture and
its structure was determined by 1H-NMR spectroscopy and (ESI–MS) as the diketene derivative
(1E,4E-1,5-bis-4-methoxyphenylpenta-1,4-dien-3-one), endorsing a reported curcumin’s
behaviour.51 On the contrary, 14 and the fluorinated-benzyloxy derivatives, obtained with similar
synthetic route as 13, showed a good chemical stability, and no degradation product was observed
during the purification procedure. The presence of the benzyloxy functions seemed to strongly
increase compound stability.
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
Blood-Brain-Barrier (BBB) Permeation
One of the main obstacles for the treatment of the diseases of the CNS is drug’s penetration into
BBB at therapeutic concentrations. The BBB is a complex interface between blood and the CNS
that strictly controls exchanges between the two compartments.52 This barrier is composed by
endothelial cells with tight junctions that protect the brain from endogenous materials which could
damage brain tissues.53 The majority of CNS drugs enter the brain by transcellular passive
diffusion, due to the tight junction structure and limited transport pathways. In early drug discovery
stage, evaluation of ADME (Absorption, Distribution, Metabolism, Excretion) properties is of
crucial importance to reduce attrition in the development process. Parallel Artificial Membrane
Permeability Assay (PAMPA), is a high throughput technique developed to predict passive
permeability through biological membranes. As curcumin and its metabolites were reported to have
poor BBB permeability,54 it seemed to be significant to evaluate the new derivatives for their impact
12
ACS Paragon Plus Environment

Page 13 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
on BBB crossing. Thus, the capacity of the selected candidates (2-6, and 8) to penetrate into the
CNS was examined through the PAMPA-BBB method described by Di et al,55 using curcumin (1b)
6
7
8
9
6
1
as standard. The in vitro permeability (Pe) values were determined, expressed as 10− cm·s− , and
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
reported in Table 3. In the light of the guidelines previously established for the prediction of the
6
1
BBB permeation,56 compounds with a Pe value superior than 4.22 10− cm·s− could be considered
as capable of crossing the BBB and classified as CNS +; a border-line profile was reported as CNS
+ / CNS -. The results indicate that compounds 2, 3, 4 and 6 may pass the BBB, while 5 and 8
showed a border-line behavior.
Neuroprotection
The neuroprotective potential was also investigated for some selected compounds by evaluating:
i) antioxidant properties, ii) induction of the detoxifying enzyme NAD(P)H: quinone
oxidoreductase 1 (NQO1), iii) cell viability. Curcumin (1b) was employed as lead compound.
i) Antioxidant properties. Oxidative stress has been recognized as a common pathological
feature in AD. It is substantially produced by reactive radical species, among others reactive oxygen
species (ROS), together with loss of function of many antioxidant defense enzymes that cause
imbalance between the formation of cellular oxidants and the antioxidative processes.57 Compounds
able to inactivate these entities could offer promises for a valuable therapeutic efficacy. The ability
of the selected curcumin-based derivatives 2, 3, 8, together with 1b to protect from oxidative stress
was studied on T67 cells, by evaluating their ROS scavenging ability after exposure to 100 μM tert-
butyl hydroperoxide (TBH) at the fixed concentration of 10 µM (Figure 5). The results showed that
2 maintained a moderate scavenging activity, slightly lower than curcumin (P = 0.005 for curcumin
vs. TBH and P = 0.08 for 2 vs. TBH), while 3 and 8, although displaying the same mean value,
show a higher standard deviation that makes them not statistically significant. The lack of
scavenging activity observed for 3 suggested that this property could not only be ascribed to the
presence of the 3-OCH3, 4-OH-aryl function. This hypothesis was further confirmed as compounds
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 41
1
2
3
4
5
6
4-7 were quite ineffective when subjected to the same evaluation (See Figure S5 of Supporting
Information).
7
8
9
ii) NQO1 induction. Human NQO1, a flavoprotein constitutively expressed in several tissues
and organs, among which the brain, exerts antioxidant effect by catalyzing the reduction of
quinones to hydroquinones and by scavenging superoxide molecules.58 In AD, as a result of a
defense mechanism against an increased oxidative stress, the levels of this phase 2 cytoprotective
enzyme were found to be remarkably increased. Interestingly, an elevated expression of this
enzyme has also been proposed as a first indicator of the pathology.59 Its induction may correlate
with a protective role by allowing to delay the progression of the malady. As curcumin itself was
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
demonstrated to effectively induce NQO1,60 a set of representative derivatives (2-8) were selected
to be tested on NQO1 induction potential, by using 1b as lead compound. As shown in Figure 6, the
tested compounds proved to be NQO1 inducers. In particular, 8 and 4 were as active as curcumin,
while higher activities were observed for 2, 3 and 6. Notably, this effect has recently been
demonstrated to slow the Aβ-induced neurodegeneration in mice.61
iii) Neuronal Cell Toxicity. Cytotoxic effects, in several in vitro or in vivo models, have been
reported for curcumin.62 The neurotoxic potential was thus examined for some selected compounds,
namely 1b, 2-8. Cells were exposed to compound concentrations ranging from 0 to 20 µM for a 24
h period of time and cell viability was measured by the MTT test, with 1b as lead compound. As
shown in Figure 7, in the experimental conditions, increasing the concentration of the tested
compound, no remarkable decrease of cell viability was observed. In particular, 4 and 7 showed
moderate toxicity at 20 μM concentration, comparable to that of 1b, that for 7 was dose-dependent.
The remaining compounds did not show any apparent cytotoxic effects at these concentrations. The
potential toxicity of 5, 4 and 6 was assayed only at the concentration of 10 µM after 24 hours of
incubation and the results confirmed the absence of any apparent cytotoxic effect.
14
ACS Paragon Plus Environment

Page 15 of 41
Journal of Medicinal Chemistry
1
2
3
Conclusion
4
5
6
7
8
The complex pathological mechanisms of AD directed attention to polypharmacological
strategies and multi-target drugs that, acting on several targets involved in the disease, could offer
promises for obtaining improved effectiveness compared to single-target drugs. In this scenario,
BACE-1 and GSK-3β emerged as AD validated targets. The curcumin pharmacophore, in its β-
keto-enol form, encompassing some structural features suitable for modulation of BACE-1 and
GSK-3β, was rationally envisaged as a potential dualistic inhibitor of both targets. Efforts addressed
to the design and synthesis of a small library of curcumin-based analogues allowed to obtain
valuable probes to explore the chemical space of these structurally unrelated enzymes. Overall, the
enol tautomeric form demonstrated to be an important feature for achieving good inhibitory potency
and chemical stability. In summary, derivative 8 showed a remarkable gap of potency (two order of
magnitude) between its inhibition of BACE-1 and GSK-3β. It indeed turned out to be a single-target
inhibitor of BACE-1 but it could very well represent a promising starting point for the development
of optimized dual modulators. Compounds 3, 9 and 10, likewise characterized by higher potencies
on BACE-1 with respect to GSK-3β (sub-micromolar and low-micromolar IC50 values,
respectively), showed a reduced deviation from the dualistic profile. Interestingly, a number of
derivatives proved to modulate BACE-1 and GSK-3β enzymes with quite comparable potencies in
the micromolar range. A mild dualistic profile was shown for 4 and 5, whereas 2, 6, and 7 proved to
be well-balanced low-micromolar inhibitors of both enzymes, a crucial achievement in the multi-
target drug discovery context. Furthermore, among this last subset, 2, 3, and 6 proved to induce
NQO1 enzyme, thus offering promises to exert a neuroprotective activity by counteracting the
oxidative stress, a common feature of AD. In particular, 2, thanks to its weak antioxidant effect,
accompanied by the absence of evident neurotoxic effects and by favourable pharmacokinetic
behaviour, clearly emerged as promising AD-modifying drug candidate, and development of
analogues derived from 2 will certainly be pursued.
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
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 41
1
2
3
4
In the light of these findings, several curcumin-based analogues were identified in this study as
5
6
7
8
valuable lead compounds that deserve attention to be optimized with different approaches, all aimed
at obtaining effective and dualistic BACE-1 / GSK-3β inhibitors, endowed with good
neuroprotective and pharmacokinetic properties.
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
EXPERIMENTAL SECTION
Chemistry. General Procedures. Starting materials, unless otherwise specified in the
Experimental Section, were used as high-grade commercial products. Solvents were of analytical
grade. Melting points were determined in open glass capillaries, using a Büchi apparatus and are
uncorrected. 1H-NMR and 13C-NMR spectra were recorded on Varian Gemini and chemical shifts
are reported as parts per million (ppm  value) relative to the peak for tetramethylsilane (TMS) as
internal standard. Standard abbreviations indicating spin multiplicities are given as follows: s
(singlet), d (doublet), t (triplet), br (broad), q (quartet) or m (multiplet). Mass spectra were recorded
on a Waters ZQ 4000 apparatus operating in electrospray mode (ES). Chromatographic separations
were performed on silica gel columns using the flash method (Kieselgel 40, 0.040-0.063 mm,
Merck). Reactions were followed by thin layer chromatography (TLC) on precoated silica gel plates
(Merck Silica Gel 60 F254) and then visualized with a UV lamp. Satisfactory elemental analyses
were obtained for all new compounds, confirming >95% purity. Compounds were named following
IUPAC rules as applied by Beilstein-Institute AutoNom (version 2.1), a PC-integrated software
package for systematic names in organic chemistry.
Pabon reaction: General Synthetic Procedure for Compounds 2-8, 17, 18. To a stirred
solution of pentane-2,4-dione (1.00 mmol) in EtOAc (1.0 mL), B2O3 (1.0 molar equiv) was added,
and the suspension was stirred for 30 min at 80 °C before addition of a solution of the appropriate
aldehyde/s, (0.9 molar equiv for monoaryl or 1.8 molar equiv for bi-aryl curcumin derivatives), tri-
n-butyl borate (2.0 molar equiv for monoaryl or 4.0 molar equiv for bi-aryl curcumin derivatives),
16
ACS Paragon Plus Environment

Page 17 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in EtOAc (0.5 mL). The reaction mixture was stirred at 80 °C for 30 min, then a solution of n-
butylammine (0.4 molar equiv in 1.0 mL of EtOAc) was added over a period of 15 min. The
mixture was heated to 80 °C for 8 h and then, after cooling to r.t., it was acidified with 0.5 N HCl
(30 mL) and then stirred at 80 °C for 30 min. The organic phase was separated and the aqueous
layer was extracted with EtOAc (3 x 10 mL). The combined organic layers were sequentially
washed with saturated aqueous NaHCO3 and brine, dried over Na2SO4, filtered, and concentrated
under reduced pressure. The crude residue was purified by flash column chromatography using a
mixture of petroleum ether / ethyl acetate (PE / EtOAc) as eluent, followed by crystallization from
suitable solvent.
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
With this procedure, the following compounds were obtained:
3Z,5E-4-hydroxy-6-(4-hydroxy-3-methoxyphenyl)hexa-3,5-dien-2-one (17). Reaction of
pentane-2,4-dione (1.00 g, 10.00 mmol), B2O3 (0.70 g, 10.00 mmol), and vanillin (1.37 g, 9.00
mmol), in EtOAc (15.0 mL), gave a crude product that was purified by flash chromatography (PE /
EtOAc, 9.75:0.25), yellow powder, 55% yield, mp 144-146 °C (EtOH). 1H NMR (400 MHz,
CDCl3):  2.16 (s, 3H, CH3), 3.94 (s, 3H, OCH3), 5.40 (br, 1H, OH), 5.63 (s, 1H, keto-enol-CH),
6.33 (d, 1H, J = 16.0 Hz, CH=CH), 6.92 (d, 1H, J = 8.0 Hz, H-5), 7.02 (d, 1H, J = 1.8 Hz, H-2),
7.09 (dd, 1H, J = 8.0 and 1.8 Hz, H-6), 7.53 (d, 1H, J = 16.0 Hz, CH=CH). Analytical data of this
intermediate are in good agreement with the literature data.63
3Z,5E-6-(4-benzyloxyphenyl)-4-hydroxyhexa-3,5-dien-2-one (18). Reaction of pentane-2,4-
dione (1.00 g, 10.00 mmol), B2O3 (0.70 g, 10.00 mmol), and 4-benzyloxybenzaldehyde (2.04 g,
9.00 mmol) gave a crude product that was purified by flash chromatography (PE / EtOAc,
9.95:0.05), yellow powder, 65% yield, mp 121-122 °C (DCM / PE). 1H NMR (400 MHz, CDCl3): 
2.18 (s, 3H, CH3), 5.20 (s, 2H, OCH2), 5.61 (s, H, keto-enol-CH), 6.33 (d, 1H, J = 16.0 Hz,
CH=CH), 6.94 (d, 2H, J = 8.2 Hz, Ar), 7.40-7.45 (m, 5H, Bn), 7.53 (d, 2H, J = 8.2 Hz, Ar), 7.53 (d,
1H, J = 16.0 Hz, CH=CH).
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 41
1
2
3
4
5
6
7
8
9
1E,4Z,6E-1-(4-benzyloxyphenyl)-5-hydroxy-7-(4-hydroxy-3-methoxyphenyl)hepta-1,4,6-
trien-3-one (2). Reaction of 17 (1.17 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-
benzyloxybenzaldehyde (0.95 g, 4.50 mmol), gave a crude product that was purified by flash
chromatography (PE / EtOAc, 9:1), orange-yellow powder, 44% yield, mp 169-170 °C (EtOH). 1H-
NMR (400 MHz, CDCl3): δ 3.96 (s, 3H, OCH3), 5.19 (s, 2H, OCH2), 5.76 (s, 1H, keto-enol-CH),
6.45 (d, 2H, J = 16.0 Hz, CH=CH), 6.94 (d, 1H, J = 8.4 Hz, H-5), 6.98 (d, 2H, J = 8.4 Hz, Ar), 7.07
(d, 1H, J = 1.8 Hz, H-2), 7.12 (dd, 1H, J = 1.8 and 8.4 Hz, H-6), 7.40-7.45 (m, 5H, Bn), 7.53 (d, 2H,
J = 8.4 Hz, Ar), 7.60 (d, 2H, J = 16.0 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 55.70, 70.44,
103.56, 111.78, 115.81, 116.37 (2C), 123.18, 123.21, 127.58, 128.17 (3C), 128.32 (2C), 128.50
(2C), 129.01, 129.33, 137.27, 140.41 (2C), 148.50, 149.52, 160.78, 186.65 (2C). ESI-MS (m/z): 451
(M + Na).
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
1E,4Z,6E-5-hydroxy-7-(4-hydroxy-3-methoxyphenyl)-1-p-tolylhepta-1,4,6-trien-3-one
(3).
Reaction of 17 (1.17 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-methylbenzaldehyde (0.53
mL, 4.50 mmol) gave a product that was purified by flash chromatography (PE / EtOAc, 9:1),
orange-yellow powder, 36% yield, mp 136-138 °C (EtOH). 1H-NMR (400 MHz, CDCl3): δ 2.38 (s,
3H, CH3), 3.95 (s, 3H, OCH3), 5.81 (s, 1H, keto-enol-CH), 6.49 (d, 1H J = 15.8 Hz, CH=CH), 6.60
(d, 1H, J = 15.8 Hz, CH=CH), 6.94 (d, 1H, J = 8.2 Hz, H-5), 7.06 (d, 1H, J = 1.8 Hz, H-2), 7.13
(dd, 1H, J = 1.8 and 8.2 Hz, H-6), 7.20 (d, 2H, J = 8.0 Hz, Ar), 7.46 (d, 2H, J = 8.0 Hz, Ar ), 7.59
(d, 1H, J = 15.8 Hz, CH=CH), 7.63 (d, 1H, J = 15.8 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3) δ
21.04, 55.70, 107.59, 112.88, 116.00, 121.46, 121.87, 126.70, 127.42 (2C), 129.27 (2C), 130.94,
134.09, 141.23 (2C), 137.91, 148.50, 150.23, 183.70 (2C). ESI-MS (m/z): 359 (M + Na).
1E,4Z,6E-7-(4-benzyloxyphenyl)-5-hydroxy-1-4-methoxyphenylhepta-1,4,6-trien-3-one (6).
Reaction of 18 (1.47 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-methoxybenzaldehyde (0.53
mL, 4.50 mmol) gave a crude product that was purified by flash chromatography (PE / EtOAc, 8:2),
light orange-yellow powder, 48% yield, mp 135-137 °C (DCM / PE). 1H-NMR (400 MHz, CDCl3):
18
ACS Paragon Plus Environment

Page 19 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
δ 3.86 (s, 3H, OCH3), 5.12 (s, 2H, OCH2), 5.80 (s, 1H, keto-enol-CH), 6.51 (d, 2H, J = 15.6 Hz,
CH=CH), 6.93 (d, 2H, J = 8.0 Hz, Ar), 7.00 (d, 2H, J = 8.0 Hz, Ar), 7.40-7.45 (m, 5H, Bn), 7.52 (d,
6
7
8
9
13
4H, J = 8.0 Hz, Ar), 7.63 (d, 2H, J = 15.6 Hz, CH=CH). C-NMR (101 MHz, CDCl3): δ 55.41,
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
70.54, 114.68, 114.45 (2C), 115.30 (2C), 122.03, 127.80 (2C), 128.10 (2C), 128.50, 128.99 (2C),
129.79, 129.93 (4C), 137.02, 140.25 (2C), 160.73, 161.01, 183.49 (2C). ESI-MS (m/z): 435 (M +
Na).
1E,4Z,6E-7-(4-benzyloxyphenyl)-5-hydroxy-1-4-hydroxyphenylhepta-1,4,6-trien-3-one (7).
Reaction of 18 (1.47 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-hydroxylbenzaldehyde (0.56
g, 4.50 mmol) gave a crude product that was purified by flash chromatography (PE / EtOAc, 9:1),
1
yellow powder, 42% yield, mp 189-191 °C (EtOH). H-NMR (400 MHz, CDCl3): δ 5.11 (s, 2H,
OCH2), 5.78 (s, 1H, keto-enol-CH), 6.49 (d, 1H, J = 16.0 Hz, CH=CH), 6.50 (d, 1H, J = 15.6 Hz,
CH=CH), 6.85 (d, 2H, J = 8.0 Hz, Ar), 7.00 (d, 2H, J = 8.8 Hz, Ar), 7.40-7.45 (m, 5H, Bn), 7.47 (d,
2H, J = 8.0 Hz, Ar), 7.51 (d, 2H, J = 8.8 Hz, Ar), 7.61 (d, 1H, J = 15.6 Hz, CH=CH), 7.62 (d, 1H, J
= 15.6 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 70.70, 114.81, 115.70 (2C), 116.53 (2C),
122.57, 127.98 (2C), 128.21 (2C), 128.29, 129.04 (2C), 129.89, 130.06 (4C), 136.87, 140.74 (2C),
160.80, 161.00, 183.58 (2C). ESI-MS (m/z): 421 (M + Na).
1E,4Z,6E-5-hydroxy-1,7-bis-4-hydroxyphenylhepta-1,4,6-trien-3-one (4). Reaction of
pentane-2,4-dione (1.00 g, 10.00 mmol), B2O3 (0.70 g, 10.00 mmol) and 4-hydroxybenzaldehyde
(2.24 g, 18.00 mmol) gave a crude product that was purified by flash chromatography (PE / EtOAc,
1
7:3), red-orange powder, 88% yield, mp 228-230 °C (EtOH). H-NMR (400 MHz, acetone-d6): δ
5.99 (s, 1H, keto-enol-CH), 6.67 (d, 2H, J = 15.8 Hz, CH=CH), 6.91 (d, 4H, J = 8.6 Hz, Ar), 7.57
(d, 4H, J = 8.6 Hz, Ar), 7.61 (d, 2H, J = 15.8 Hz, CH=CH). 1H-NMR (400 MHz, DMSO-d6): δ 6.22
(s, 1H, keto-enol-CH), 6.82 (d, 2H, J = 16.4 Hz, CH=CH), 6.98 (d, 4H, J = 8.4 Hz, Ar), 7.25 (d, 4H,
J = 8.4 Hz, Ar), 7.35 (br, 1H, OH), 7.62 (d, 2H, J = 16.4 Hz, CH=CH), 9.70 (br, 2H, OH).64 13C-
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 41
1
2
3
4
5
6
NMR (101 MHz, acetone-d6): δ 114.00, 116.50 (4C), 122.86, 127.93 (2C), 129.71, 130.01 (4C),
141.08 (2C), 161.07 (2C), 183.50 (2C). ESI-MS (m/z): 331 (M + Na).
7
8
9
1E,4Z,6E-5-hydroxy-1,7-bis-4-methoxyphenylhepta-1,4,6-trien-3-one (5). Reaction of
pentane-2,4-dione (1.00 g, 10.00 mmol), B2O3 (0.70 g, 10.00 mmol), and 4-methoxybenzaldehyde
(2.19 mL, 18.00 mmol) gave a crude product that was purified by flash chromatography (PE /
EtOAc, 9:1) yellow powder, 87% yield, mp 110-112 °C (DCM / PE). 1H-NMR (400 MHz, CDCl3):
δ 3.86 (s, 6H, OCH3), 5.79 (s, 1H, keto-enol-CH), 6.51 (d, 2H, J = 15.6 Hz, CH=CH), 6.93 (d, 4H, J
= 7.2 Hz, Ar), 7.52 (d, 4H, J = 7.2 Hz, Ar), 7.63 (d, 2H, J = 15.6 Hz, CH=CH). 1H-NMR (400
MHz, DMSO-d6): δ 3.80 (s, 6H, OCH3), 6.09 (s, 1H, keto-enol-CH), 6.79 (d, 2H, J = 15.6 Hz,
CH=CH), 7.00 (d, 4H, J = 8.4 Hz, Ar), 7.59 (d, 2H, J = 15.6 Hz, CH=CH), 7.68 (d, 4H, J = 8.4 Hz,
Ar). 13C-NMR (101 MHz, CDCl3): δ 55.38 (2C), 113.75, 114.37 (4C), 121.80, 127.80 (2C), 129.51,
129.75 (4C), 140.09 (2C), 161.25 (2C), 183.33 (2C). 13C-NMR (101 MHz, DMSO-d6): δ 55.38
(2C), 101.43, 114.52 (4C), 121.82 (2C), 127.33 (2C), 130.18 (4C), 140.05 (2C), 161.09 (2C),
183.12 (2C). ESI-MS (m/z): 359 (M + Na).
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
1E,4Z,6E-1,7-bis-4-benzyloxyphenyl-5-hydroxyhepta-1,4,6-trien-3-one (8). Reaction of
pentane-2,4-dione (1.00 g, 10.00 mmol), B2O3 (0.70 g, 10.00 mmol), and 4-benzyloxybenzaldehyde
(3.80 g, 18.00 mmol) gave a crude product that was purified by flash chromatography (PE / EtOAc,
1
9:1), yellow powder, 87% yield, mp 161-162 °C (DCM / PE). H-NMR (400 MHz, CDCl3): δ 5.11
(s, 4H, OCH2), 5.78 (s, 1H, keto-enol-CH), 6.50 (d, 2H, J = 16.0 Hz, CH=CH), 6.99 (d, 4H, J = 8.4
Hz, Ar), 7.40-7.44 (m, 10H, Bn), 7.51 (d, 4H, J = 8.4 Hz, Ar), 7.62 (d, 2H, J = 16.0 Hz, CH=CH).
1H-NMR (400 MHz, DMSO-d6): δ 5.17 (s, 4H, OCH2), 6.09 (s, 1H, keto-enol-CH), 6.79 (d, 2H, J =
16.4 Hz, CH=CH), 7.08 (d, 4H, J = 8.4 Hz, Ar), 7.40-7.43 (m, 10H, Bn), 7.59 (d, 4H, J = 8.4 Hz,
Ar), 7.68 (d, 2H, J = 15.6 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 70.44 (2C), 115.61 (5C),
122.27, 127.79 (2C), 127.81 (2C), 128.39 (2C), 128.50 (2C), 128.98 (2C), 129.00 (2C), 130.07,
130.12 (4C), 136.80 (2C), 140.41 (2C), 160.78 (2C), 183.67 (2C). ESI-MS (m/z): 511 (M + Na).
20
ACS Paragon Plus Environment

Page 21 of 41
Journal of Medicinal Chemistry
1
2
3
1E,4Z,6E-1,7-bis-4-(4-fluorobenzyloxyphenyl)-5-hydroxyhepta-1,4,6-trien-3-one
(11).
4
5
6
Reaction of pentane-2,4-dione (0.50 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-(4-
7
8
9
fluorobenzyloxybenzaldehyde 19 (2.07 g, 9.00 mmol) gave a crude product that was purified by
1
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
flash chromatography (PE / EtOAc, 9:1), yellow solid, 31% yield, mp 210-212 °C. H-NMR (400
MHz, CDCl3): δ 5.06 (s, 4H, OCH2), 5.78 (s, 1H, keto-enol-CH), 6.50 (d, 2H, J = 15.6 Hz,
CH=CH), 6.98 (d, 4H, J = 8.8 Hz, Ar), 7.09-7.13 (m, 4H, Ar), 7.41-7.44 (m, 4H, Ar), 7.51 (d, 4H, J
= 8.8 Hz, Ar), 7.62 (d, 2H, J = 15.6 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 70.89 (2C),
114.50, 115.01 (d, 4C, J = 27.3 Hz), 115.72 (4C), 122.97, 128.89, 129.42 (4C), 129.47 (2C), 129.71
(d, 4C, J = 8.1 Hz), 131.94 (d, 2C, J = 4.0 Hz), 140.82 (2C), 159.67 (2C), 162.49 (d, 2C, J = 265.6
Hz), 183.76 (2C). ESI-MS (m/z): 547 (M + Na).
1E,4Z,6E-1,7-bis-4-(3-fluorobenzyloxyphenyl)-5-hydroxyhepta-1,4,6-trien-3-one
(12).
Reaction of pentane-2,4-dione (0.50 g, 5.00 mmol), B2O3 (0.35 g, 5.00 mmol), and 4-(3-
fluorobenzyloxybenzaldehyde 20 (2.07 g, 9.00 mmol) gave a crude product, that was purified by
1
flash chromatography (PE / EtOAc, 9:1), 35% yield, yellow solid, mp 160-162 °C. H-NMR (400
MHz, CDCl3): δ 5.10 (s, 4H, OCH2), 5.78 (s, 1H, keto-enol-CH), 6.51 (d, 2H, J = 15.6 Hz,
CH=CH), 6.98 (d, 4H, J = 8.8 Hz, Ar), 7.01-7.05 (m, 2H, Ar), 7.16-7.18 (m, 2H, Ar), 7.20-7.22 (m,
2H, Ar), 7.35-7.37 (m, 2H, Ar), 7.51 (d, 4H, J = 8.8 Hz, Ar), 7.62 (d, 2H, J = 15.6 Hz, CH=CH).
13C-NMR (101 MHz, CDCl3): δ 72.62 (2C), 114.16 (d, 2C, J = 28.3 Hz), 115.07 (d, 2C, J = 26.3
Hz), 115.27 (5C), 123.31, 123.63 (d, 2C, J = 4.0 Hz), 127.81 (4C), 128.99 (3C), 129.28 (d, 2C, J =
8.1 Hz), 140.80 (d, 2C, J = 6.9 Hz), 143.82 (2C), 160.59 (2C), 163.07 (d, 2C, J = 264.6 Hz), 183.67
(2C). ESI-MS (m/z): 547 (M + Na).
Williamson reaction: General Procedure for the Synthesis of 8-10, 13-16.
To a solution of phenol-derivative (1.00 mmol) in acetone (10.0 mL), anhydrous K2CO3 (2.2
molar equiv) and the appropriate alkyl or aryl halide (2.2 molar equiv) were added. The reaction
mixture was heated to 80 °C and reaction progress was monitored by TLC. Upon reaction
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 41
1
2
3
4
5
6
7
8
9
completion, the mixture was hot filtered and the solvent was evaporated under reduced pressure.
The two tautomers obtained were effectively isolated from the resulting crude mixture by column
chromatography over silica gel and using a mixture of PE / EtOAc as eluent. The desired final
compounds were further purified by fractionated crystallization from DCM / PE. In particular,
regarding the order of elution of the tautomeric mixtures, the diketo tautomer proved to elute first.
With this procedure, the following compounds were obtained:
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
1E,4Z,6E-1,7-bis-4-benzyloxyphenyl-5-hydroxyhepta-1,4,6-trien-3-one (8) and 1E,6E-1,7-
bis-4-benzyloxyphenylhepta-1,6-diene-3,5-dione (14). Reaction of 4 (0.50 g, 1.62 mmol), K2CO3
(0.49 g, 3.56 mmol), and benzyl bromide (0.47 mL, 3.56 mmol) gave a tautomeric mixture that was
purified by flash chromatography (PE / EtOAc, 9:1). 8 (Rf: 0.12), experimental characterization
matched with that of the same analogue obtained by the Pabon reaction (see above). 14 (Rf: 0.15),
1
32% yield, dark-yellow powder, mp 145-147 °C. H-NMR (400 MHz, CDCl3): δ 3.96 (s, 2H, β-
diketo-CH2), 5.08 (s, 4H, OCH2), 6.87 (d, 2H, J = 16.0 Hz, CH=CH), 6.94 (d, 4H, J = 8.4 Hz, Ar),
7.35-7.39 (m, 10H, Bn), 7.52 (d, 4H, J = 8.4 Hz, Ar), 7.73 (d, 2H, J = 16.0 Hz, CH=CH). 13C-NMR
(101 MHz, CDCl3): δ 55.73, 70.44 (2C), 115.60 (4C), 122.27 (2C), 127.85 (4C), 128.40 (2C),
128.50 (2C), 129.20 (4C), 130.00 (4C), 136.91 (2C), 140.50 (2C), 160.80 (2C), 196.37 (2C). ESI-
MS (m/z): 511 (M + Na).
1E,4Z,6E-1,7-bis-4-(2,4-difluorobenzyloxy)phenyl)-5-hydroxyhepta-1,4,6-trien-3-one
(9)
and 1E,6E-1,7-bis-4-(2,4-difluorobenzyloxyphenyl)hepta-1,6-diene-3,5-dione (15). Reaction of
4 (0.50 g, 1.60 mmol) and 2,4-difluorobenzyl bromide (0.48 mL, 3.52 mmol) gave a crude product
that was purified by flash chromatography (PE / EtOAc, 9.5:0.5). 9 (Rf: 0.13), yellow powder, 32%
yield, 174-176 °C. 1H-NMR (400 MHz, CDCl3): δ 5.12 (s, 4H, OCH2), 5.79 (s, 1H, keto-enol-CH),
6.51 (d, 2H, J = 16.0 Hz, CH=CH), 6.84-6.95 (m, 4H, Ar), 6.99 (d, 4H, J = 8.4 Hz, Ar), 7.46-7.48
(m, 2H, Ar), 7.52 (d, 4H, J = 8.4 Hz, Ar), 7.61 (d, 2H, J = 16.0 Hz, CH=CH). 13C-NMR (101 MHz,
CDCl3): δ 63.50 (2C), 104.00 (t, 2C, J = 25.6 Hz), 111.63 (dd, 2C, J = 21.4 and 3.8 Hz), 114.98
22
ACS Paragon Plus Environment

Page 23 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(4C), 115.17, 119.13 (dd, 2C, J = 14.6 and 4.0 Hz), 123.31, 128.48 (4C), 128.50 (2C), 130.52,
130.85 (dd, 2C, J = 9.8 and 5.3 Hz), 140.81 (2C), 160.74 (dd, 2C, J = 238.4 and 12.0 Hz), 163.23
(dd, 2C, J = 237.9 and 12.0 Hz), 163.39 (2C), 183.65 (2C). ESI-MS (m/z): 583 (M + Na). 15 (Rf:
1
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
0.16), yellow powder, 28% yield, 153-155 °C. H-NMR (400 MHz, CDCl3): δ 3.98 (s, 2H, β-
diketo-CH2), 5.11 (s, 4H, OCH2), 6.82 (d, 2H, J = 15.6 Hz, CH=CH), 6.88-6.92 (m, 4H, Ar), 6.96
(d, 4H, J = 8.4 Hz, Ar), 7.10-7.12 (m, 2H, Ar), 7.46 (d, 4H, J = 8.8 Hz, Ar), 7.75 (d, 2H, J = 15.6
Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 55.43, 63.49 (2C), 104.04 (t, 2C, J = 25.7 Hz),
111.62 (dd, 2C, J = 21.9 and 4.0 Hz), 116.37 (4C), 119.20 (dd, 2C, J = 15.1 and 4.5 Hz), 126.43
(2C), 128.59 (4C), 129.02 (2C), 130.90 (dd, 2C, J = 10.6 and 5.4 Hz), 143.07 (2C), 160.60 (dd, 2C,
J = 238.7 and 12.2 Hz), 160.80 (2C), 163.03 (dd, 2C, J = 238.6 and 12.2 Hz), 196.30 (2C). ESI-MS
(m/z): 583 (M + Na).
1E,4Z,6E-1,7-bis-4-(3,5-difluorobenzyloxyphenyl)-5-hydroxyhepta-1,4,6-trien-3-one
(10)
and 1E,6E-1,7-bis-4-(3,5-difluorobenzyloxyphenyl)hepta-1,6-diene-3,5-dione (16). Reaction of
4 (0.50 g, 1.60 mmol) and 3,5-difluorobenzyl bromide (0.48 mL, 3.52 mmol) gave a crude product
that was purified by flash chromatography (PE / EtOAc, 9.5:0.5). 10 (Rf: 0.11), 60%, yield, mp
1
174-176 °C. H-NMR (400 MHz, CDCl3): δ 5.09 (s, 4H, OCH2), 5.80 (s, 1H, keto-enol-CH), 6.52
(d, 2H, J = 15.6 Hz, CH=CH), 6.56-6.67 (m, 2H, Ar), 6.73-6.81 (m, 2H, Ar), 6.93-6.99 (m, 2H, Ar),
6.97 (d, 4H, J = 8.8 Hz, Ar), 7.52 (d, 4H, J = 8.8 Hz, Ar), 7.63 (d, 2H, J = 15.6 Hz, CH=CH). 13C-
NMR (101 MHz, CDCl3): δ 71.64 (2C), 102.80 (t, 2C, J = 27.7 Hz), 112.18 (dd, 4C, J = 27.3 and
4.0 Hz), 114.99, 115.14 (4C), 122.98, 128.87 (4C), 129.05, 130.10 (2C), 141.60 (t, 2C, J = 7.2 Hz),
143.81 (2C), 159.97 (2C), 164.26 (dd, 4C, J = 261.8 and 6.9 Hz), 183.50 (2C) ESI-MS (m/z): 583
(M + Na). 16 (Rf: 0.13) yellow powder, 28% yield, mp 152-154 °C. 1H-NMR (400 MHz, CDCl3): δ
3.94 (s, 2H, diketo-CH2), 5.07 (s, 4H, OCH2), 6.60-6.72 (m, 2H, Ar), 6.75 (d, 2H, J = 15.2 Hz,
CH=CH), 6.79-6.84 (m, 2H, Ar), 6.93 (d, 4H, J = 8.4 Hz, Ar), 6.93-6.98 (m, 2H, Ar), 7.43 (d, 4H, J
= 8.8 Hz, Ar), 7.75 (d, 2H, J = 15.2 Hz, CH=CH). 13C-NMR (101 MHz, CDCl3): δ 55.84, 71.73
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 41
1
2
3
4
5
6
7
8
9
(2C), 102.98 (t, 2C, J = 27.7 Hz), 112.20 (dd, 4C, J = 27.0 and 4.0 Hz), 115.30 (4C), 127.13 (2C),
129.48 (4C), 129.50 (2C), 140.57 (t, 2C, J = 6.8 Hz), 142.41 (2C), 160.59 (2C), 164.20 (dd, 4C, J =
261.3 and 6.4 Hz) 195.06 (2C). ESI-MS (m/z): 583 (M + Na).
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
1E,6E-1,7-bis-4-methoxyphenylhepta-1,6-diene-3,5-dione (13). Reaction of 4 (0.15 g, 0.49
mmol), K2CO3 (0.15 g, 1.08 mmol), and iodomethane (0.07 mL, 1.08 mmol) gave the crude product
that was purified by flash chromatography (PE / EtOAc, 9.5:0.5) from which 13 was obtained as
1
predominant component, orange-yellow powder, 44% yield, mp 103-105 °C. H-NMR (400 MHz,
CDCl3): δ 3.48 (s, 2H, diketo-CH2), 3.86 (s, 6H, OCH3), 6.93 (d, 4H, J = 8.8 Hz, Ar), 7.01 (d, 2H, J
13
= 14.8 Hz, CH=CH), 7.55 (d, 4H, J = 6.8 Hz, Ar), 7.72 (d, 2H, J = 14.8 Hz, CH=CH). C-NMR
(101 MHz, CDCl3) δ 55.38 (2C), 55.43, 114.37 (4C), 121.80 (2C), 128.24 (2C), 129.75 (4C),
141.20 (2C), 161.37 (2C), 196.57 (2C). ESI-MS (m/z): 359 (M + Na).
4-(4-fluorobenzyloxy)benzaldehyde (19). Reaction of 4-hydroxybenzaldehyde (1.24 g, 10.00
mmol) and 4-fluorobenzylbromide (1.49 mL, 12.00 mmol) gave a crude product that was purified
by flash chromatography (PE / EtOAc, 9.5:0.5), white solid, 88% yield, mp 190-192 °C. 1H-NMR
(400 MHz, CDCl3): δ 5.08 (s, 2H, OCH2), 6.99-7.00 (m, 2H, Ar), 7.02 (d, 2H, J = 8.8 Hz, Ar), 7.10-
7.14 (m, 2H, Ar), 7.82 (d, 2H, J = 8.8 Hz, Ar), 9.80 (s, 1H, CHO).
4-(3-fluorobenzyloxy)benzaldehyde (20). Reaction of 4-hydroxybenzaldehyde (1.24 g, 10.00
mmol) and 3-fluorobenzylbromide (1.49 mL, 12.00 mmol) gave a crude product that was purified
by flash chromatography (PE / EtOAc, 9.5:0.5), white solid, 85% yield, mp 183-185 °C. 1H-NMR
(400 MHz, CDCl3): δ 5.10 (s, 2H, OCH2), 6.95 (d, 2H, J = 8.8 Hz, Ar), 6.93-6.99 (m, 1H, Ar), 7.00-
7.07 (m, 2H, Ar), 7.23-7.31 (m, 1H, Ar), 7.84 (d, 2H, J = 8.8 Hz, Ar), 9.80 (s, 1H, CHO).
Chemical Stability Studies.
The tested derivative were dissolved in DMSO (0.50 mg / mL) and the pH of the solution was
adjusted to 7.4 and 9.2 using 50 mM phosphate and 50 mM borate buffer, respectively. The
24
ACS Paragon Plus Environment

Page 25 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
obtained solutions were then maintained at room temperature and at 70 °C (oven) for three days and
analyzed by RP-HPLC under the following conditions: the stationary phase was a phenyl-hexyl
column (Xselect CSH by Waters) 4.6 x 150 mm (particle size 3.5 µm); the mobile phase was a
mixture of ACN-0.1% formic acid, 70-30 (v / v) at the flow rate of 1 mL / min. The detection
wavelength was 254 nm.
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
BACE-1 inhibition.
FRET inhibition studies were performed using the following procedures: 5 μL of test compound
(or DMSO) were pre-incubated with 175 μL of BACE-1 (17.2 nM, final concentration) in 20 mM
sodium acetate pH 4.5 containing CHAPS (0.1% w / v) for 1 hour at room temperature. M-2420 (3
μM, final concentration) was then added and left to react for 15 min at 37 °C. The fluorescence
signal was read at λem = 405 nm (λexc = 320 nm). DMSO concentration in the final mixture was
maintained below 5% (v / v) to guarantee no significant loss of enzyme activity. Fluorescence
intensities with and without inhibitors were registered and compared. The percent inhibition due to
the presence of test compounds was calculated. The background signal was measured in control
wells containing all the reagents, except hrBACE-1, and subtracted. The % inhibition due to the
presence of test compound was calculated by the following expression: 100-(IFi / IFo x 100) where
IFi and IFo are the fluorescence intensities obtained for hrBACE-1 in the presence and in the
absence of inhibitor, respectively.43 The linear regression parameters were determined and the IC50
interpolated (GraphPad Prism 4.0, GraphPad Software Inc.). To demonstrate inhibition of BACE-1
activity a peptido-mimetic inhibitor (β-secretase inhibitor IV, Calbiochem) was serially diluted into
the reactions’ wells (IC50 = 20 nM).
Inhibition of GSK-3β.
Human recombinant GSK-3β was purchased from Millipore (Millipore Iberica S.A.U.). The
prephosphorylated polypeptide substrate was purchased from Millipore (Millipore Iberica SAU).
Kinase-Glo Luminescent Kinase Assay was obtained from Promega (Promega Biotech Iberica, SL).
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 41
1
2
3
4
ATP and all other reagents were from Sigma-Aldrich (St. Louis, MO). Assay buffer contained 50
5
6
7
8
mM HEPES (pH = 7.5), 1 mM EDTA, 1 mM EGTA, and 15 mM magnesium acetate. The method
of Baki et al.46 was followed to analyze the inhibition of GSK-3β. Kinase-Glo assays were
performed in assay buffer using white 96-well plates. In a typical assay, 10 μL (10 μM) of tested
compound (dissolved in DMSO at 1 mM concentration and diluted in advance in assay buffer to the
desired concentration) and 10 μL (20 ng) of enzyme were added to each well followed by 20 μL of
assay buffer containing 25 μM substrate and 1 μM ATP. The final DMSO concentration in the
reaction mixture did not exceed 1%. After a 30 min incubation at 30 °C, the enzymatic reaction was
stopped with 40 μL of Kinase-Glo reagent. Glow-type luminescence was recorded after 10 min
using a Fluoroskan Ascent multimode reader. The activity is proportional to the difference of the
total and consumed ATP. The inhibitory activities were calculated on the basis of maximal kinase
and luciferase activities measured in the absence of inhibitor and in the presence of reference
compound inhibitor (SB415826,65 supplier, IC50 = 54 nM) at total inhibition concentration,
respectively. The linear regression parameters were determined and the IC50 extrapolated
(GraphPad Prism 4.0, GraphPad Software Inc.).
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
Covalent docking simulation on GSK-3β
The curcumin-optimized binding pocket from the best scoring complex of the SCARE docking
procedure was used to perform a covalent docking of 5. A thia-Michael reaction occurring between
Cys199 and the α,β-unsaturated carbonyl function of 5 was simulated by means of ICM3.7 standard
covalent docking protocol.66
Cell culture.
The T67 human glioma cell line was derived by Lauro et al. from a World Health Organization
(WHO) Grade III gemistocytic astrocytoma. T67 cells were cultured in Dulbecco’s modified
Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 UI / mL
penicillin, 100 µg / mL streptomycin in a 5% CO2 atmosphere at 37 °C, with saturating humidity.
26
ACS Paragon Plus Environment

Page 27 of 41
Journal of Medicinal Chemistry
1
2
3
Determination of antioxidant activity in T67 cells.
4
5
6
7
8
To evaluate the antioxidant activity of the compounds, T67 cells were seeded in 24-well plates at
1 x 105 cells / well. After 24 hours, the cells were washed with PBS and treated for 24 hours with
the compounds at 10 µM concentration. The antioxidant activity of the compounds was evaluated
after 30 minutes of incubation with 10 µM fluorescent probe (2’,7’-dichlorofluorescein diacetate,
DCFH-DA) in DMEM, by measuring the intracellular ROS formation evoked by 30 minutes
exposure of T67 cells to 100 µM tert-butyl hydroperoxide (TBH) in PBS. The fluorescence increase
of the cells from each well was measured (λexc = 485 nm; λem = 535 nm) with a
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
spectrofluorometer (Wallac Victor multilabel S9 counter, Perkin-Elmer Inc., Boston, MA). Data are
reported as the mean ± standard deviation of at least three independent experiments.
Determination of cell viability (Cytotoxicity Assays)
Cytotoxicity of compounds was estimated using an MTT assay. T67 cells were seeded with
complete DMEM in 24-well plates and cultured overnight. Cells were washed with PBS and
incubated for 24 hours at 37 °C in humidified atmosphere, 5% CO2, with compounds at different
concentrations dissolved in DMEM. After this time, cells were carefully washed with PBS and
incubated for 3 h with 500 µM thiazolyl blue tetrazolium bromide (MTT) dissolved in DMEM.
Then, cells were washed with PBS and lysed with 200 μL of dimethyl sulfoxide. The absorbance of
each well was measured at 595 nm using a microplate reader (Victor2 1420 multilabel counter).
Preparation of cell homogenate
T67 cells were treated for 24 hours with [10 µM] of compounds and [2.5 µM] of 4-
methylsulfinylbutyl isothiocyanate (sulforaphane) as positive control. After this time, cells were
collected and resuspended in ice-cold 50 mM potassium phosphate buffer, pH 7.4, containing 2 mM
EDTA and 0.1% Triton X-100.
Assay for NQO1 induction.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 41
1
2
3
4
5
6
7
8
9
Cellular NQO1 activity was measured according to the procedures described previously.67
Briefly, the reaction mix contained 50 mM Tris–HCl, pH 7.5, 0.08% Triton X-100, 0.25 mM
NADH, and 80 µM of 2,6-dichloroindophenol (DCIP) in the presence or absence of 10 µM
dicumarol. To an assay cuvette, 0.980 mL of the reaction mix was added and the reaction was
started by adding 20 µL of cell homogenate and the reduction of DCIP was monitored at 600 nm,
30 °C for 3 min. The dicumarol-inhibitable NQO1 activity was calculated using the extinction
coefficient of 21.0 mM-1 cm-1 and expressed as nmol of DCIP reduced per min / mg of cellular
protein determined by Lowry method.68
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
Determination of the Chemical Stability
Vehicle (DMSO), 5 and 13 (0.5 mg / mL) at pH 7.4 and 9.2 (adjusted with aqueous buffers) were
incubated for three days at 70 °C (oven).
28
ACS Paragon Plus Environment

Page 29 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
Table 1. Inhibition of BACE-1 and GSK-3β enzymatic activities by the curcumin-based analogues
as β-keto-enol tautomer (2-12) and reference compound curcumin (1b).
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
IC50 (µM)a
± SEM
comp
R
OH
R’
OCH3
OCH3
H
R”
BACE-1
0.97 ± 0.43
0.14 ± 0.03
GSK-3β
0.90 ± 0.38
2.09 ± 0.51
8.39 ± 1.59
0.53 ± 0.27
2.78 ± 0.44
2.01 ± 0.71
2.49 ± 0.82
9.63 ± 0.21
2
3
4
5
6
7
8
9
OH
CH3
OH
2.54 ± 0.02
17 ± 3c
OH
OCH3
H
OCH3
OCH3
OH
1.65 ± 0.01
2.28 ± 0.64
2.69 ± 1.01
0.04 ± 0.01
0.40 ± 0.06
H
H
H
H
H
0.39 ± 0.34
8.30 ± 0.54
10
H
H
1.04 ± 0.34
1.08 ± 0.66
12.81 ± 0.14
11
12
16.99 ± 2.68
17.95 ± 1.03
n.i.b
Curc
(1b)
343 ± 45c
29
ACS Paragon Plus Environment