Hydroxy-substituted trans-cinnamoyl derivatives as multifunctional tools in the context of Alzheimer's disease

Article abstract of DOI:10.1016/j.ejmech.2017.07.058

Alzheimer's disease (AD) is a multifactorial pathology that requires multifaceted agents able to address its peculiar nature. In recent years, a plethora of proteins and biochemical pathways has been proposed as possible targets to counteract neurotoxicity. Although the complex scenario is not completely elucidated, close relationships are emerging among some of these actors. In particular, increasing evidence has shown that aggregation of amyloid beta (Aβ), glycogen synthase kinase 3β (GSK-3β) and oxidative stress are strictly interconnected and their concomitant modulation may have a positive and synergic effect in contrasting AD-related impairments. We designed compound 3 which demonstrated the ability to inhibit both GSK-3β (IC₅₀ = 24.36 ± 0.01 μM) and Aβ₄₂ self-aggregation (IC₅₀ = 9.0 ± 1.4 μM), to chelate copper (II) and to act as exceptionally strong radical scavenger (k_inh = 6.8 ± 0.5 · 10⁵ M^?1s^?1) even in phosphate buffer at pH 7.4 (k_inh = 3.2 ± 0.5 · 10⁵ M^?1s^?1). Importantly, compound 3 showed high-predicted blood-brain barrier permeability, did not exert any significant cytotoxic effects in immature cortical neurons up to 50 μM and showed neuroprotective properties at micromolar concentration against toxic insult induced by glutamate.

Full text of DOI:10.1016/j.ejmech.2017.07.058

Accepted Manuscript
Hydroxy-substituted trans-cinnamoyl derivatives as multifunctional tools in the context
of Alzheimer's disease
Angela De Simone, Manuela Bartolini, Andrea Baschieri, Kim Y.P. Apperley, Huan
Huan Chen, Melissa Guardigni, Serena Montanari, Tereza Kobrlova, Ondrej Soukup,
Luca Valgimigli, Vincenza Andrisano, Jeffrey W. Keillor, Manuela Basso, Andrea
Milelli
PII:
S0223-5234(17)30580-9
10.1016/j.ejmech.2017.07.058
EJMECH 9621
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 15 March 2017
Revised Date: 29 May 2017
Accepted Date: 24 July 2017
Please cite this article as: A. De Simone, M. Bartolini, A. Baschieri, K.Y.P. Apperley, H.H. Chen,
M. Guardigni, S. Montanari, T. Kobrlova, O. Soukup, L. Valgimigli, V. Andrisano, J.W. Keillor, M.
Basso, A. Milelli, Hydroxy-substituted trans-cinnamoyl derivatives as multifunctional tools in the
context of Alzheimer's disease, European Journal of Medicinal Chemistry (2017), doi: 10.1016/
j.ejmech.2017.07.058.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Hydroxy-substituted Trans-cinnamoyl Derivatives as Multifunctional Tools in the Context of
Alzheimer’s Disease
Angela De Simone,a Manuela Bartolini,b Andrea Baschieri,c Kim Y.P. Apperley,d Huan Huan
Chen,a Melissa Guardigni,a Serena Montanari,a Tereza Kobrlova,e Ondrej Soukup,e Luca
Valgimigli,c Vincenza Andrisano,a Jeffrey W. Keillor,d Manuela Basso,f Andrea Milellia*
a
Department for Life Quality Studies, Alma Mater Studiorum-University of Bologna, Corso
d’Augusto 237, 47921 Rimini, Italy
b
Department of Pharmacy and Biotechnology, Alma Mater Studiorum-University of Bologna, Via
Belmeloro 6, 40126 Bologna, Italy
c Department of Chemistry "G. Ciamician", Alma Mater Studiorum-University of Bologna, Via S.
Giacomo 11, 40126 Bologna, Italy
d
Department of Chemistry and Biomolecular Sciences, University of Ottawa, 10 Marie-Curie,
Ottawa, ON, Canada K1N 6N5
e
Biomedical Research Center, University Hospital, Sokolska 581, 500 05 Hradec Kralove Czech
Republic
f Laboratory of Transcriptional Neurobiology, Centre for Integrative Biology, University of Trento,
Via Sommarive 9, 38123 Trento, Italy
1

ACCEPTED MANUSCRIPT
Corresponding Author. Dr. Andrea Milelli, Phone: Tel. +390541434610, Fax
+390541434608. E-mail: andrea.milelli3@unibo.it.
Graphical abstract
GSK-3β
Aβ42 self
aggregation
Ethyl Ferulate
ROS
Abstract
Alzheimer’s disease (AD) is a multifactorial pathology that requires multifaceted agents able to
address its peculiar nature. In recent years, a plethora of proteins and biochemical pathways has
been proposed as possible targets to counteract neurotoxicity. Although the complex scenario is not
completely elucidated, close relationships are emerging among some of these actors. In particular,
increasing evidence has shown that aggregation of amyloid beta (Aβ), glycogen synthase kinase 3β
(GSK-3β) and oxidative stress are strictly interconnected and their concomitant modulation may
have a positive and synergic effect in contrasting AD-related impairments. We designed compound
3 which demonstrated the ability to inhibit both GSK-3β (IC50 = 24.36 ± 0.01 µM) and Aβ42 self-
aggregation (IC50 = 9.0 ± 1.4 µM), to chelate copper (II) and, in addition, to act as exceptionally
strong radical scavenger (kinh = 3.2 ± 0.5 · 105 M-1s-1) even in phosphate buffer at pH 7.4 (kinh = 3.2
± 0.5 · 105 M-1s-1). Importantly, compound 3 showed high-predicted blood-brain barrier
2

ACCEPTED MANUSCRIPT
permeability, did not exert any significant cytotoxic effects in immature cortical neurons up to 50
µM and showed neuroprotective properties at micromolar concentration against toxic insult induced
by glutamate.
1. Introduction
Multifactorial diseases, such as cancer and neurological disorders, are driven by dysregulation of
different but interconnected biochemical pathways. The complexity of these pathologies makes
them hardly addressable by single-target molecules. In response, during the last two decades the
scientific community has focused on a “multitarget drug design” approach, which may represent a
major breakthrough in drug discovery.[1, 2] The cornerstone of this design strategy is the
assumption that a multifactorial disease iscould be better contrasted by a chemical entity
characterized by a similar complexity, called a MultiTarget Ligand (MTL). Indeed, the
simultaneous modulation of multiple targets by a MTL should more significantly impact the course
of disease progression.
One of the most active area of application of this strategy has been against the search for an
effective treatment for Alzheimer’s disease (AD).[3] AD is a multifactorial and heterogeneous
pathology characterized by the aberrant processing of two key players, namely amyloid-β peptide
(Aβ), that leads to the formation and accumulation of senile plaques, characterized by aggregates of
Aβ,[4] and tau-protein, whose hyperphosphorylation leads to the formation of neurofibrillary
tangles (NFTs).[5] These aggregates, together with other pathogenic factors, such as oxidative
stress, inflammation and excitotoxicity, trigger a cascade of interrelated pathways leading,
ultimately, to neuronal cell death.[6]
Worldwide costs of AD are enormous ($604 billion in 2010 [7]) and are estimated to further
increase if no effective treatment will be found in the close future. Notwithstanding the urgent need
for new therapies and in spite of massive efforts and investments, effective treatments are still
unavailable.[8] Indeed, currently approved drugs are not effective in preventing, halting, or
reversing AD the pathology, but only temporally ameliorating cognitive impairment by enhancing
the cholinergic tone (acetylcholinesterase inhibitors-AChEIs) or counteracting the excitotoxicity
excitotoxicity exerted by glutamate (antagonist at the NMDA receptor) in the brain of AD patients.
A survey of the drug candidates currently in clinical trials for AD [9] shows that several approaches
addressing all critical pathological players in AD, i.e., amyloid- and tau-related toxicity,
neuroinflammation, oxidative stress and neurodegeneration, are under investigation, without a clear
higher beneficial advantage of one approach over the other.
3

ACCEPTED MANUSCRIPT
The majority of the MTL strategies so far investigated have relied on coupling the scaffold of
known AChEIs, such a s tacrine, with a second fragment displaying additional biological activities,
such as inhibition of Aβ aggregation or Aβ formation, scavenging activity towards reactive oxygen
species (ROS), metal chelating properties and neuroprotection activity.[10-12]
Recently, iIn the search for more effective compounds, recently,, new MTLs based on non-AChEI
scaffolds have been proposed. Bolognesi and co-workers have reported a new series of MTLs based
on the triazinone scaffold as dual β-secretase (BACE-1)/glycogen synthase kinase 3β (GSK-3β)
inhibitors.[13, 14] These agents inhibit enzymes that belong to the two main toxic pathways of AD,
i.e., Aβ and tau-protein pathways, and display promising neuroprotective and neurogenic activities.
These results further highlight the benefit of simultaneously targeting distinct points of the
neurotoxic cascades.
BACE-1 is an aspartyl protease that is overexpressed in the brain of AD patients and cleaves the
amyloid precursor protein (APP) leading, ultimately, to the formation and accumulation of Aβ
peptide.[15] Therefore, treatments with BACE-1 inhibitors are envisioned as long-term therapy to
limit Aβ production after removing existing amyloid deposits. On the other hand, GSK-3β is
responsible for tau hyperphosphorylation which triggers the formation of NFTs. [16] Furthermore,
GSK-3β has been suggested as a possible link between Aβ and tau. [17]
Several BACE-1 inhibitors are currently in different stages of clinical development.[9] However,
therapeutic tuning of BACE-1 activity is a challenging task as demonstrated by the recent failure of
verubecestat.[18] The complexity of such approach raises from the fact that APP is not the only
physiological target of BACE-1 and it plays key roles in the myelination of neurons, in learning,
memory, and in synaptic plasticity. [19] Furthermore, inhibition of BACE-2, a close homologous
enzyme of BACE-1, is responsible for side effects in pancreas, skin and brain.[20]
Other than reducing Aβ generation through BACE-1 inhibition, neurotoxicity induced by Aβ
aggregates can be counteracted by other approaches, such as increasing Aβ clearance, preventing its
aggregation or interfering with the boosting effect on Aβ toxicity exerted by the dyshomeostasis of
biometals in the brain. [21] Indeed, although the debate on the role of amyloid in the onset and
progression of AD is still open, drug candidates directed toward Aβ have shown to be promising
and have reached the last phases of clinical trials (fourteen drug candidates directed towards Aβ
were in phase III clinical trials in June 2016).[9] Among these, aducanumab (Biogen), a human
recombinant monoclonal antibody that selectively targets Aβ aggregates,[22] was granted Fast
Track designation by the U.S. Food and Drug Administration (FDA) in 2016, underlining the
4

ACCEPTED MANUSCRIPT
expected positive outcome from such an approach. These findings have renewed the interest of the
scientific community for the amyloid hypothesis.[23]
Finally, several antioxidants with a multitarget mechanisms of action are currently investigated in
clinical trials.[9] Indeed, it is well known that neurodegeneration processes in the central nervous
system (CNS) are often initiated or enhanced by oxidative stress and that a critical link exists, since
it promotes Aβ toxicity, and vice versa, and triggers the amyloidogenic pathway. Furthermore,
oxidative stress was reported to affect the tau protein phosphorylation status and significantly
increase GSK-3β activity. Supporting these findings, the antioxidant compound melatonin was
shown to decrease GSK-3β activity and tau hyperphosphorylation in vitro and in vivo.
Finally, compelling evidence has shown that neurodegenerative processes in the central nervous
system (CNS) are often initiated or enhanced by an increased oxidative stress, which precedes the
appearance of the hallmarks of the disease.[24-26] Furthermore, oxidative stress was reported to
affect tau protein phosphorylation [27] and to significantly increase GSK-3β activity.[28, 29]
Supporting these findings, the antioxidant compound melatonin was shown to decrease GSK-3β
activity and tau hyperphosphorylation in vitro and in vivo.[30] These observations laid down the
basis for the clinical investigation of several antioxidants with a multitarget mechanisms of
action.[9]
On these premises, the aim of the present investigation is the discovery of a pharmacological tool
able to block the neurodegenerative cascade by operating at critical points of the two major
pathogenic mechanisms, i.e., targeting both Aβ self-aggregation and GSK-3β activity, and to disrupt
the vicious circle involving oxidative stress, Aβ toxicity and tau hyperphosphorylation (Figure 1).
To this purpose, we focused our attention on ferulic acid ethyl ester (ethyl-4-hydroxy-3-
methoxycinnamic acid), a natural antioxidant which exerts protective effects on neurons against
Aβ-induced toxicity by modulating the oxidative stress both directly and indirectly through the
induction of protective genes.[31] Furthermore, the chemical structure of ferulic acid ethyl ester
also offered the possibility to be easily modified in order to broaden its biological activity. Indeed,
the trans-cinnamoyl scaffold represents a privileged structure and plenty of compounds based on
this structure have been evaluated against a plethora of biological targets.[32-34]
Five trans-cinnamoyl derivatives (1-5) were designed to i) act as antioxidants thanks to the
substituted aromatic moiety deriving from natural products, such as ferulic and caffeic acids and
ethyl ferulate, ii) interact with GSK-3β through the α,β-unsaturated system[35] and iii) counteract
Aβ42 self-aggregation thanks to the substituted aromatic moiety and the presence of an extended
planar conjugate system.[36] The introduction of a 3-pyridynyl 3-pyridyl ring was envisaged to
5

ACCEPTED MANUSCRIPT
have a double role. Indeed, its extended conjugate system could allow a better interaction with Aβ42
growing fibrils and establish additional interactions with GSK-3β.
Figure 1. Design strategy leading to derivatives 1-5.
To confirm the multifunctional activity profile envisaged for the designed compounds, derivatives
1-5 were evaluated for their ability to inhibit GSK-3β and Aβ42 self-aggregation and to counteract
reactive oxygen radical (ROS) formation.
Furthermore, for For the most interesting compound, the ability to chelate copper, a biometal
involved in AD neurotoxicity, was evaluated together with the neuroprotective activity against
glutamate-induced toxicity in cortical neurons. Furthermore, due to the structural similarity to well-
known inhibitors of transglutaminase 2 (Tg2)[37] and the role of this enzyme in AD pathogenesis,
the inhibitory activity towards Tg2 was also determined. Indeed, Tg2 is an enzyme that catalyzes
the post translational modifications of proteins whose expression and activities are increased in
brain areas affected by AD.[38] Tg2 was suggested to be able to initiate Aβ aggregation and induce
the formation of high-molecular weight tau aggregates. Therefore, inhibition of Tg2 might
implement the overall beneficial activity of a new therapeutic for AD treatment. Lastly, since
activity should be exerted in the CNS, for representative compounds the prediction of the blood-
brain barrier (BBB) permeability was investigated by BBB-PAMPA assay.
2. Chemistry
The synthesis of the target compounds is reported in Scheme 1. Briefly, the hydroxyl group(s) of
the appropriate aldehydes were protected as tetrahydropyranyl ethers and the resulting compounds
(6-10) were then subjected to a classic base-catalyzed aldol reaction with 3-acetylpyridine, leading
to α,β-unsaturated ketones 11-15. Acidic hydrolysis of the aldol adducts 11-15 led to the
corresponding final compounds 1-5.
6

ACCEPTED MANUSCRIPT
Scheme 1. a) 3,4-Dihydro-2H-pyran, pyridinium p-toluenesulfonate, dichloromethane, rt, 12 h; b)
3-acetylpyridine, 2.5 M aq. NaOH, rt, 12 h; 38-60 yields; c) p-Toluenesulfonic acid, methanol, rt,
12 h; 41-72 yields.
2. Results and Discussion
2.1 Inhibitory activity toward GSK-3β
Based on the expected activity profile, compounds 1-5 were evaluated for their ability to inhibit
GSK-3β, in comparison with ethyl ferulate, using a luminescence assay. [39] Compounds were first
screened at a single concentration, namely 50 µM, and for the most active derivative the IC50 value
was measured. As reported in Table 1, although the introduction of a pyridynyl 3-pyridyl group
induced an increase in the inhibitory activity compared to ethyl ferulate, all compounds, with the
exception of 3, show poor inhibitory potency with values ranging from 25.2 to 43.5% ([I] = 50
µM). Surprisingly, Compound 3, characterized by a 2,3-dihydroxy substitution on the aromatic ring,
turned out to be a good inhibitor of GSK-3β with an IC50 value of 24.36 ± 0.01 µM. The structural
requirements for a good inhibition seem to be quite strict. Indeed, the substitution of an hydroxyl
group with a methoxy or ethoxy group, as in compounds 1 and 2, or the presence of a single
hydroxy group on the aromatic ring, as in 4 and 5, led to compounds with a weak inhibitory
activity.
The inhibitory potency of compound 3 lies in the micromolar range, which is in line with the
activity of other GSK-3β-based multipotent compounds.[14] Evidence from in vivo studies suggests
7

ACCEPTED MANUSCRIPT
that inhibition at micromolar concentration would be adequate to produce therapeutic effects in the
brain without affecting GSK-3β activity in peripheral tissues.[40]
2.2 Inhibition of Aβ42 self-aggregation
The anti-aggregating activity toward Aβ42 of compounds 1-5 was determined by a thioflavin (ThT)-
based fluorometric assay.[41] The ThT-based assay is among the most widely used methods to
screen small molecules as inhibitors of amyloid aggregation and fibrillization. In such an assay,
emission at 490 nm is assumed to be directly proportional to the quantity of formed amyloid fibrils.
To exclude any positive or negative interference in the fluorescence signal due to quenching or
displacement phenomena, fluorescence emission intensities of preformed Aβ42 fibrils (in a 1.5-µM
ThT solution) in the absence and presence of inhibitor were compared. No significant interference
was observed under the experimental conditions (∆IF ≤10% for the highest tested concentration).
Similarly to what was observed for the inhibition of GSK-3β, all the compounds, with the exception
of compound 3, turned out to be poor inhibitors with inhibition ranging between 5.8 and 21.5%
when tested at a 1:1 ratio with Aβ42. Differently to the other compounds of the series, compound 3
proved to be a potent inhibitor of Aβ42 self-aggregation with an IC50 of 9.0 ± 1.4 µM. Therefore, the
substitution pattern of the aromatic ring strongly influenced the ability to interfere with Aβ42
aggregation, since all modification of the 3,4-dihydroxyl-derivative proved to be ineffective. The
good anti-aggregating activity of 3 is in agreement with the well-known ability of polyphenols to
modulate Aβ aggregation.[42] Furthermore, these findings are partially in line with the report from
Reinke et al. concerning a structure-activity relationships (SAR) study on curcumin derivatives.[36]
Indeed, the activity profile of compounds 1-5 clearly stated that the nature of the substitution of the
aryl group is more critical than the size of the planar aromatic system for a productive interaction
with Aβ.[36] Compound 3 showed a higher inhibitory potency than that of the parent compounds
ferulic acid, 3-hydroxycinnamic acid (IC50 > 100 µM for both compounds)[36] and caffeic acid
(32.3 ± 3.7 % inhibition at 20 µM).[43] For the sake of completeness it must be pointed out the
different ionization status of the latter compounds in comparison with derivatives 1-5 in the
experiment conditions that might influence the inhibitory activity.
On the basis of data available in the literature, compound 3 can be ranked as a strong inhibitor of
Aβ42 self-aggregation, with potency very close to that of the well-known multipotent compound
bis(7)-tacrine (IC50 = 8.4 ± 1.4 µM)[44] and slightly more potent than catechol derivative recently
proposed by Simoni et al. [(S)-allyl (E)-3-(3,4-dihydroxyphenyl)prop-2-enethioate, IC50 = 12.5 ±
0.9 µM).[45]
8

ACCEPTED MANUSCRIPT
Table 1. Inhibition of GSK-3β, Aβ42 self-aggregation, Tg2 and antioxidant activity.
antioxidant activity
in PhCl (T = 30 °C)
GSK-3βa
Aβ42 self-aggregation
Tg2 IC50
(µM)±
SEM
Compounds
% inhibition
IC50 (µM)
% inhibition
± SEMc
IC50 (µM)
± SEM
kinh/M-1s-1d
ne
± SEMb
± SEM
37.18 ± 0.10
39.55 ± 0.60
n.d.
n.d.
17.5 ± 1.7
18.8 ± 0.5
n.d.
n.d.
7.1 ± 0.2 ó 103
6.1 ± 0.3 ó 103
6.8 ± 0.5 ó 105
3.2 ± 0.5 ó 105f
5.7± 0.3 ó 103
1.5 ± 0.2 ó 103
2.1 ± 0.1
2.1 ± 0.1
2.6 ± 0.3
4.5 ± 0.2
2.0 ± 0.2
1.9 ± 0.2
688 ± 105
484 ± 15
1
2
3
100
24.36 ± 0.01
91.5 ± 0.5
9.0 ± 1.4
1001 ± 35
25.18 ± 0.01
43.55 ± 0.01
n.d.
n.d.
21.5 ± 5.8
5.8 ± 5.3
n.d.
n.d.
> 2000
266 ± 22
n.d.
4
5
Ethyl
ferulate
21.38 ± 0.06
n.d.
n.d.
n.d.
4.6 ± 0.2 ó 103
2
a Values are the mean of two independent measurements, each performed in duplicate.b Assay performed in presence of
c
d
50 µM inhibitor Assays were carried out in the presence of 50 µM inhibitor and 50 µM Aβ42. Rate constant for
reaction with alkylperoxyl radicals. e Stoichiometric factor = number of peroxyl radicals trapped by one molecule of
antioxidant. f Experiment performed in 0.1M aqueous phosphate buffer (pH = 7.4), THF 3.1M, [AAPH] 25 mM, T =
30°C. nd stands for not determined; SEM stands for standard error of the mean.
2.3 Copper chelating properties
Strong relationships exist between Aβ and biometals, and changes in the transition metals, zinc,
copper and iron, have shown to impact the molecular mechanisms of the disease.[46] High
concentration of metals have been found in amyloid plaques.[47] It is also reported that copper is
able to contribute to the formation of neurotoxic ROS[48] and accelerate the formation of Aβ
oligomers.[49] Thus, agents targeting such metals have been proposed as an alternative strategy for
the treatment of AD[46] and extensive preclinical and clinical data underlie the rational of such
approach.
Based on its interesting activity profile towards GSK-3β and Aβ42, compound 3 was selected for
further characterization. In particular, the chemical structure of compound 3 makes it a potential
chelating agent. Thus, it was investigated for its ability to complex Cu2+ ions by means of UV-vis
differential spectroscopy. The UV-vis spectrum of compound 3 in methanol showed band maxima
at 244 nm (ε = 10136 M-1 cm-1), 266 nm (ε = 11013 M-1 cm-1) and 374 nm (ε = 19087 M-1 cm-1) as
detailed in the Supporting Information. Upon addition of CuCl2, absorbance at those wavelengths
was reduced and a new band appeared with maximum at 460 nm, which increased with the metal
concentration (Figure 2). Spectral changes were most accurately monitored from the differential
spectra, which were obtained by numerically subtracting the spectra of ligand and metal (at the
corresponding molar concentration) from the spectrum of the mixture (Figure 2B). Since they were
dependent on the relative concentration of metal/ligand and showed saturability, it was possible to
9

ACCEPTED MANUSCRIPT
use UV–Vis spectroscopy to determine the stoichiometry of complex formation. When increasing
concentrations of CuCl2 were added to a 7.5 µM solution of compound 3 (in methanol) the signal at
460 nm, attributable to complex formation, progressively increased up to the addition of 0.5
equivalents of Cu2+. No further increase of the absorbance was observed from 0.5 to 2 equivalents
of Cu2+. This spectral variation paired a corresponding decrease in the absorbance at 374 nm
(Figure 2C), indicating a preferred stoichiometry for the 3/Cu2+ complex of 2:1, as expected for
Cu2+ complexes with catechol.[50] Analysis of Figure 2C indicates that, even at low micromolar
concentrations, Cu2+ ions are quantitatively chelated by compound 3. While this very high affinity
prevented a quantitative assessment of the binding constant, it confirmed the expected relevance of
this interaction under biomimetic settings.
Figure 2. (A) UV–Vis (200–600 nm) absorption spectra of 3 (7.5 µM) in methanol after addition of
ascending amounts of CuCl2 (0.30–3–3.75 µM); (B) differential spectra of 3-Cu2+ complex
formation obtained by numerical subtraction of CuCl2 and 3 spectra from the above spectra, at the
corresponding concentrations, and (C) plot of the differential absorbance at 460 and 374 nm as a
function of the equivalents of Cu2+ added to the solution.
In the light of the confirmed chelating properties, compound 5 might also prevent biometal induced-
amyloid aggregation and attenuate the neurotoxic cascade triggered by the amyloid-metal complex.
Because of the potentially relevant therapeutic implications these aspects deserve further
investigation.
2.4 Antioxidant activity
Because of the strict connection between Aβ, tau, biometals and oxidative stress, compounds 1-5
were also evaluated for their ability to counteract ROS. The antioxidant activity was investigated by
measuring the kinetics of oxygen consumption during the controlled inhibited autoxidation of a
standard substrate, which is currently the best-established method to obtain mechanistic insights and
10

ACCEPTED MANUSCRIPT
accurate quantitative data to draw SARs[51]. With this method, in the absence of antioxidants,
oxygen consumption during the autoxidation (forcibly initiated at a rate Rì by an azo-compound) is
constant. Addition of micromolar concentrations of the test antioxidant would produce either a neat
inhibited period (until the antioxidant is consumed) or a decrease in the slope of oxygen-uptake
plots, whose analysis (e.g. by eqs. 1-2) affords the absolute rate constant kinh for the reaction
between the antioxidant and the oxidative chain-carrying peroxyl radicals, and the so-called
stoichiometric factor, n, i.e. the number of peroxyl radicals trapped by one molecule of antioxidant.
Representative plots recorded for compounds 1-5 are shown in Figure 3. Because of similar
structural motif and electronic properties, ethyl ferulate, a representative structure among
polyphenolic antioxidants,[52] was investigated for comparison.
Figure 3. Oxygen consumption during the autoxidation of cumene (3.6 M) initiated by AIBN (0.05
M) in PhCl at 30°C without inhibitors (red) or in the presence of compounds 1-5 and reference
compound ethyl ferulate (24 µM).
All investigated compounds were able to effectively inhibit the autoxidation of cumene in apolar
organic solution (PhCl), each trapping 2 peroxyl radicals, as expected for a typical phenolic
antioxidant, as summarized in Table 1. Rate constants nicely reflected the substitution pattern in the
phenolic ring. Compound 5, bearing only the electron withdrawing (EW) 3-pyridyl-oxopropenyl
group in the meta position, showed the lowest reactivity, while isomeric compound 4 had almost 4-
11

ACCEPTED MANUSCRIPT
fold enhanced reactivity due to stabilization of the intermediate product phenoxyl radical by
delocalization of the unpaired electron in the oxopropenyl chain. Indeed, while EW substituents are
known to increase the O-H bond dissociation enthalpy (BDE) of phenols, thereby decreasing their
reactivity, extension of the conjugated π-systems works in the opposite direction.[53] The
compensating behavior of the two factors is clearly visible upon considering the reactivity of
reference ethyl ferulate, which is identical to that previously measured for guaiacol (2-
methoxyphenol, kinh= 4.7×105 M-1s-1 [52]). In ethyl ferulate, as well as in compounds 1 and 2,
reactivity is also affected by the presence of the ortho-alkoxy substituent, which, again, acts by a
combination of (partially) compensating effects: the net electron donating (ED) character of the RO
substituent in ortho position would lower the phenolic O-H BDE, but it is almost compensated by
the introduction of an intramolecular H-bond and an increase in steric hindrance around the reactive
site, which impairs the reactivity.[54, 55] As a result, compound 1 was approximately 20% more
effective than compound 4, while the advantage was lost with compound 2 due to higher steric
hindrance.
The EW 3-pyridyl moiety has been extensively investigated in recent years to afford more stable,
less toxic and more effective antioxidants.[53] Its contribution is visible in compounds 1 and 2,
which were respectively 33% and 54% more effective that reference ethyl ferulate in trapping
peroxyl radicals. Most interesting, however, was the contribution of the chosen structural motif 3-
pyridyl-oxopropenyl group in the antioxidant behaviour of compound 3, whose reactivity surpassed
any other investigated compound by two orders of magnitude, being nearly 150-fold more effective
than reference ethyl ferulate. Such reactivity may be due to a combination of factors, summarized in
Scheme 2. The catechol ring is a privileged structure for antioxidant activity among phenols, since
the intramolecular H-bonding between the neighboring OH groups impairs the reactivity of one
(acting as H-bond donor) while it greatly enhances the reactivity of the other OH (acting as H-bond
acceptor), due to the strengthening of the H-bond as the reaction with peroxyl radicals proceeds
with the formation of the corresponding H-bonded semiquinone radical.[53]
Delocalization of the unpaired electron in the 3-pyridyloxopropenyl moiety further stabilizes the
radical, which makes compound 3 over 2-fold more reactive than structurally related caffeic acid
(kinh = 2.9×105 M-1s-1),[56] surpassing the antioxidant performance of quercetin (kinh = 5×105 M-1s-1)
and matching that of 5-hydroxytyrosol, the well-established antioxidant found in olive (kinh
8×105), measured under identical conditions.[57]
=
12

ACCEPTED MANUSCRIPT
Scheme 2. Reaction of compound 3 with peroxyl radicals in apolar organic solvents.
Since compound 3 showed the highest antioxidant activity among tested compounds, its reactivity
was further investigated in aqueous phosphate buffer solution, using tetrahydrofuran as a model
oxidation substrate.[58] At variance with other phenolic and polyphenolic antioxidants that
experience major impairment of their antioxidant performance in water due to H-bonding to the
solvent– which typically reduces the kinh of up to two orders of magnitude - compound 3 maintained
exceptionally high reactivity with kinh of 3.2×105 M-1s-1. By comparison with previous
investigations with unsubstituted catechol and other phenols, a mechanistic explanation can be
suggested as depicted in Scheme 3.
Scheme 3. Reaction mechanism of compound 3 with peroxyl radicals in water.
The EW character of the side chain increases the acidity of the catechol to the point where it causes
a shift in the reaction mechanism, from a formal H-atom transfer, i.e. the concerted electron-proton
transfer (EPT) to the peroxyl radical, to a dual mechanism in which the proton is transferred to the
solvent (or to the buffer) and one electron is transferred to the peroxyl radical. The duality of the
mechanism is related to the relative timing of the PT and ET events that can either occur stepwise
13

ACCEPTED MANUSCRIPT
(PT-ET mechanism) or concertedly but to different acceptors in a so-called, multi-site EPT (PT to
the solvent and ET to the peroxyl radical). With both mechanistic possibilities, H-bonding to the
solvent is not expected to slow down the reaction to any major extent,[58] as observed
experimentally.
Another aspect of the antioxidant behaviour of compound 3 that merits discussion is its
exceptionally high stoichiometric factor in water (n = 4.5 as opposed to typical n = 2). Again this is
attributed to the electronic properties of the side chain enabling rapid Michael-type nucleophilic
addition of water to the ortho-quinone formed upon trapping the first two peroxyl radicals.
Enolization to a polyhydroxylated aromatic ring allows subsequent rapid reaction with additional
peroxyl radicals, as illustrated in Scheme 4.
Scheme 4. Proposed mechanism for extended stoichiometry of antioxidant activity of compound 3
in water.
Beside expressing antioxidant behaviour, potentially, reducing antioxidants (including ascorbate)
could also express a paradoxical pro-oxidant behaviour under some settings; specifically, as in the
case of autoxidations initiated by Fenton-type chemistry in the presence of transition metal ions.
[51] This would be linked to their ability to reduce (i.e. recycle) metals from their higher oxidation
state (e.g. Cu2+ to Cu+), making them available for reaction with peroxides. [51] To investigate this
possibility for compound 3, we performed a series of experiments in water where autoxidation was
initiated by the redox couple Cu+/Cu2+ in the presence of hydrogen peroxide. In no case we
observed a pro-oxidant behaviour for compound 3. A representative experiment is shown in Figure
S3. Clearly, this does not exclude that compound 3 can express pro-oxidant chemistry to some
14

ACCEPTED MANUSCRIPT
extent; however this is overwhelmed by the antioxidant behaviour, that prevails under any tested
experimental settings.
2.5 Inhibitory activity toward Tranglutaminase 2
Some cinnamoyl-pyridines have been reported as Tg2 inhibitors.[37] Tg2 is an enzyme that
catalyzes the post translational modifications of proteins whose expression and activities are
increased in brain areas affected by AD.[38] In particular, Tg2 is observed in senile plaques[59].
Furthermore, NFT[60], Aβ and tau, either in phosphorylated and non-phosphorylated forms, are
substrates of Tg2-catalyzed modifications.[61, 62] Experimental evidence suggests that Tg2 may
initiate the aggregation process of Aβ and induces the formation of high-molecular weight tau
aggregates, which are more resistant to enzymatic degradation. Furthermore, Tg2 inhibitors were
shown to counteract Aβ-induced toxicity in the SH-SY5Y cell line.[63] Therefore, compounds 1-5
were evaluated for their ability to inhibit Tg2. Results in Table 1 show that these compounds act as
low to moderate inhibitors of Tg2, depending on the substitution pattern of the hydroxyl/alkoxy
groups on the aromatic ring. Compound 5 was found to be the most potent Tg2 inhibitor within this
series, although its potency is roughly an order of magnitude lower than that reported for the
inhibition of guinea pig liver TGase by similar azachalcones.[37]
2.6 Cell toxicity and neuroprotection
Glutamate is a major excitatory neurotransmitter in mammalian CNS. Excessive glutamate
releasing overactivates its receptors causing calcium dyshomeostasis, increasing the production of
ROS and initiating a cascade of intracellular events leading to neuronal degeneration. Based on its
promising in vitro profile, compound 3 was investigated for its protective effect on neurons
undergoing oxidative stress. In particular, immature cortical neurons were treated at day in vitro 1
(DIV1) with glutamate, to block the Xct- transporter and reduce the intracellular levels of
glutathione, and with increasing concentrations of compound 3.[64, 65] Viability was assessed by
MTT assay after 24 hours in three different neuronal preparations. Interestingly, compound 3 did
not show any toxic effects in this cell line up to 50 µM, resulting safe at concentrations at which it
exerts its biological activity. Furthermore, as shown in Figure 5, compound 3 was able to block
neuronal death induced by glutamate (5 mM) already at the concentration of 12.5 µM.
15

ACCEPTED MANUSCRIPT
Figure 5. Compound 3 protects cortical neurons from oxidative-stress mediated cell death.
Immature neurons (DIV1) were treated with increasing concentrations of compound 3 in presence
(black bars) and in absence (white bars) of glutamate, able at high concentration (5mM) to deplete
neurons of their major antioxidant, glutathione, and induce oxidative stress. Compound 3 blocks
glutamate toxicity already at low µM concentrations with statistical significance at 50µM. The data
were analyzed with two-way ANOVA and Bonferroni’s post hoc test; **p<0.01 and it is referred to
the significant difference between untreated and glutamate treated neurons (0µM). #p<0.05 and it
compared the neuronal viability in glutamate-treated neurons in the absence of compound 3 with
glutamate-treated neurons in the presence of compound 3 (50µM).
2.7 Blood–brain barrier penetration
Penetration across the BBB is an essential property for compounds targeting the CNS. Thus, in
order to predict passive BBB permeability of novel compounds a modification of the parallel
artificial membrane permeability assay (BBB-PAMPA) was used.[66, 67] PAMPA assay is widely
used as pre-screening tool in early drug discovery to investigate drug permeation through biological
barriers. Data obtained for compounds 1, 3 and 5 were correlated to standard drugs among which
the anti-AD drugs donepezil, rivastigmine and tacrine. For such reference drugs the Permeability
(Pe) values were previously determined using PAMPA assay in the same experimental conditions
and their CNS availability is known.[66, 68] Prediction of BBB penetration of compounds 1, 3, and
5 is summarized in Table 2. Compounds 1, 3 and 5 showed Pe values similar to those determined
for the reference AD drugs and have high probability to cross the BBB via passive diffusion [Pe (10-
6 cm s-1) > 4.0].
16

ACCEPTED MANUSCRIPT
Table 2: Prediction of blood-brain barrier penetration of drugs expressed as Pe ± SEM (n=2).
BBB penetration estimation
Compound
Pe ± SEM (*10-6 cm s-1)
CNS (+/-)
9.9 ± 0.4
5.5 ± 0.3
6.5 ± 0.4
7.3 ± 0.9
6.6 ± 0.5
5.3 ± 0.2
11.3 ± 1.6
5.6 ± 0.6
2.9 ± 0.1
2.2 ± 0.1
1.1 ± 0.2
1.0 ± 0.4
CNS (+)
CNS (+)
CNS (+)
CNS (+)
CNS (+)
CNS (+)
CNS (+)
CNS (+)
CNS (+/-)
CNS (+/-)
CNS (-)
CNS (-)
1
3
5
Donepezil
Rivastigmine
Tacrine
Testosterone
Chlorpromazine
Hydrocortisone
Piroxicam
Theophyline
Atenolol
CNS (+)’ (high BBB permeation predicted); Pe (10-6 cm s-1) > 4.0; CNS (-) (low BBB permeation predicted); Pe
(10-6 cm s-1) < 2.0; CNS (+/-) (BBB permeation uncertain); Pe (10-6 cm s-1) from 4.0 to 2.0.
3. Conclusions
AD is a multifactorial disease whose development depends on the dysregulation of several
interconnected biochemical pathways. For instance, strict connections exist between the main
neurotoxic pathways involved in AD, namely Aβ and tau, and oxidative stress is emerging as a key
modulator of the toxicity induced by both proteins. Indeed, oxidative stress promotes Aβ toxicity
and increases GSK-3β mediated tau hyperphosphorylation. Finally, biometals accumulation in the
brain may increase amyloid-related toxicity. Therefore, the development of a single molecule able
to contrast both Aβ and tau-induced toxicity, and break the vicious circle induced by oxidative
stress, could represent a promising strategy in the design of effective anti-AD agents. As a proof of
concept of the feasibility of generating a single molecule able to exert such a multiple action, we
reported the design and preliminary characterization of compound 3 which may be able to contrast
AD neurotoxicity by acting at different levels of the neurotoxic cascade. In particular, experimental
outcomes clearly indicate that compound 3 is able to modulate GSK-3β and Aβ42 self-induced
aggregation. Furthermore, compound 3 acts as copper chelator and, more interestingly, exerts a very
high ROS scavenging activity in aqueous medium. A detailed investigation of the scavenging
properties of related compounds allowed clarifying the exceptionally high scavenging activity.
17

ACCEPTED MANUSCRIPT
Considering the transversal role of oxidative stress, which seems to be the “fil rouge” of the whole
AD pathology, being cited as causing or resulting pathological hallmark in all hypothesis for AD
pathogenesis, the high antioxidant activity combined with the inhibitory actions toward GSK-3β
and Aβ42 makes 3 a very promising compound.
Finally, compound 3 showed to be safe when tested in immature cortical neurons up to 50 µM and
was able to protect these neurons from toxic stimuli induced by glutamate and was predicted to
have a good BBB permeability. In light of these considerations, compound 3 could be considered a
promising tool to investigate AD pathogenesis and may be worth for further development in the
search for an effective anti-AD drug candidate.
4. Experimental Section
4.1 Chemistry
Uncorrected melting point was taken in glass capillary tubes on a Buchi SMP-20 apparatus. The
elemental analysis was performed with Perkin Elmer elemental analyzer 2400 CHNESI. MS spectra
were recorded on Perkin Elmer 297 and Waters ZQ 4000 and Micromass Q-TOF Ultima Global. 1H
NMR and 13C NMR were recorded on Varian VRX 200 and 400 instruments. Chemical shifts are
reported in parts per million (ppm) relative to peak of tetramethylsilane (TMS) and spin
multiplicities are given as s (singlet), br s (broad singlet), d (doublet), dd (doublet of doublets) t
(triplet), td (doublet of triplets), q (quartet) or m (multiplet). Chromatographic separations were
performed on silica gel columns by flash (Kieselgel 40, 0.040 e 0.063 mm, Merck) column
chromatography. Reactions were followed by thin layer chromatography (TLC) on Merck
(0.25 mm) glass-packed pre-coated silica gel plates (60 F254) and then visualized in an iodine
chamber or with a UV lamp. Reagents and solvents have been purchased by Sigma Aldrich and
used as received unless otherwise reported. THF was distilled and stored under argon at 5°C; the
content in hydroperoxides was determined periodically by spectrophotometry at 262 nm in
isopropanol upon reaction with triphenylphosphine, and found <50 ppm (µg g-1). Cumene and
styrene were purified by double percolation through silica and activated alumina columns before
use. AIBN was recrystallized from methanol and stored at -18 °C. Phosphate buffer solution (pH
7.4) were prepared as previously described, were mixed with the desired amount of THF (3:1 by
volume) after having adjusted the pH to the desired value.[69]
18

ACCEPTED MANUSCRIPT
4.1.1 General procedure for the synthesis of compounds 1-5
To the appropriate protected aldol-adduct 11-15 in methanol, para-toluensulfonic acid para-
toluenesulfonic acid (cat.) was added and the resulting mixture was stirred at room temperature
overnight. The solvent was removed under vacuum and the obtained residue was purified through
flash chromatography leading to the unprotected final products 1-5.
4.1.1.1 (E)-3-(4-hydroxy-3-methoxyphenyl)-1-(pyridin-3-yl)prop-2-en-1-one (1)
Compound 1 was obtained from compound 11; eluting mixture dichloromethane / methanol
1
(9.8:0.2); yellow solid; 61% yield; mp = 181-183 °C; H-NMR (CDCl3, 400 MHz) δ 3.99 (s, 3H),
6.98 (d, 1H, J = 8.0 Hz), 7.15 (d, 1H, J = 2), 7.23-7.25 (dd, 1H) J = 2, 1.3 Hz), 7.27 (d, 1H, J = 2.4
Hz), 7.34 (d, 1H, J = 15.6 Hz), 7.47-7.50 (m, 1H), 7.80 (d, 1H, J = 15.6 Hz), 8.29-8.32 (dt, 1H, J =
6, 2 Hz), 8.81 (d, 1H, J = 3.6 Hz), 9.24 (brs, 1H); 13C-NMR (CDCl3, 400 MHz) δ 56.00, 110.08,
115.09, 118.88, 123.92, 126.90, 136.02, 146.50, 147.04, 149.00. 149.42, 152.67, 189.01; HMRS
(esi) m/z calcd for [M+H] 256.0974, found 256.0961.
4.1.1.2 (E)-3-(3-ethoxy-4-hydroxyphenyl)-1-(pyridin-3-yl)prop-2-en-1-one (2)
Compound 2 was obtained from compound 12; eluting mixture mobile phase dichloromethane /
1
toluene / methanol (9:0.8:0.2); yellow solid; 55% yield; mp = 179-181 °C; H-NMR (CDCl3, 400
MHz) δ 1.37 (t, 3H, J = 6.8 Hz), 4.11-4.16 (q, 2H, J = 6.8, 6.8 Hz), 6.86 (d, 1H, J = 8 Hz), 7.29 (d,
1H, J = 7.6 Hz), 7.54 (s, 1H), 7.58-7.61 (dd, 1H, J = 5.2, 2.4 Hz), 7.69-7.80 (q, 2H, J = 15.2, 12.4
Hz), 8.43 (d, 1H, J = 7.6 Hz), 8.81 (d, 1H, J = 8.8 Hz), 9.33 (s, 1H), 9.68 (s, 1H); 13C-NMR
(DMSO-d6, 100 MHz) δ 14.67, 64.01, 112.90, 115.64, 118.29, 123.86, 124.63, 126.05, 133.16,
135.78, 145.76, 147.17, 149.56, 150.30, 153.02, 188.07; HMRS (esi) m/z calcd for [M+H]
270.1130, found 270.1147.
4.1.1.3. (E)-3-(3,4-dihydroxyphenyl)-1-(pyridin-3-yl)prop-2-en-1-one (3)
Compound 3 was obtained from compound 13; eluting mixture mobile phase dichloromethane /
1
methanol (9.5:0.5); yellow solid; 58% yield; mp = 237-239 °C; H-NMR (DMSO-d6, 400 MHz) δ
6.81 (d, 1H, J = 8 Hz), 7.20-7.23 (dd, 1H, J = 8, 2 Hz), 7.28 (d, 1H, J = 2), 7.56-7.59 (m, 1Hdd, J =
12.8, 3.2 Hz), 7.64 (s, 2H), 8.40-8.43 (dt, 1H, J = 8, 2), 8.79-8.80 (dd, 1H, J = 4.8, 1.6 Hz), 9.26 (d,
1H, J = 1.6 Hz); 13C-NMR (DMSO-d6, 100 MHz) δ 115.75, 115.82, 118.20, 122.57, 123.87,
126.08, 133.23, 135.78, 145.64, 145.79, 149.12, 149.47, 153.00, 188.15; HMRS (esi) m/z calcd for
[M+H] 242.0817, found 242.0802.
4.1.1.4. (E)-3-(4-hydroxyphenyl)-1-(pyridin-3-yl)prop-2-en-1-one (4)
Compound 4 was obtained from compound 14; eluting mixture mobile phase dichloromethane /
1
methanol (9.5:0.5); yellow solid; 72% yield; mp = 213-215 °C; H-NMR (CD3OD-d4, 400 MHz) δ
6.84 (d, 1H, J = 8.4 Hz), 7.56-7.66 (m, 4H), 7.80 (d, 1H, J = 15.6 Hz), 8.45 (d, 1H, J = 8 Hz), 8.83
19

ACCEPTED MANUSCRIPT
(d, 1H, J = 3.6 Hz), 8.18 (s, 1H); 13C-NMR (DMSO-d6, 100 MHz) δ 115.84, 118.23, 123.86,
125.58, 131.30, 133.14, 135.77, 145.29, 149.49, 153.02, 160.42, 188.14; HMRS (esi) m/z calcd for
[M+H] 226.0868, found 226.0857.
4.1.1.5. (E)-3-(3-hydroxyphenyl)-1-(pyridin-3-yl)prop-2-en-1-one (5):
Compound 5 was obtained from compound 15; eluting mixture mobile phase dichloromethane /
1
methanol (9.5:0.5); yellow solid; 41% yield; mp = 209-211 °C; H-NMR (CD3OD-d4, 400 MHz) δ
6.69 (m, 1H), 6.78-6.89 (dt, 1H, J = 7.2, 2.4 Hz), 7.06-7.07 (m, 1H), 7.15-7.20 (m, 2H), 7.53-7.55
(dd, 1H, J = 5.0, 3.2 Hz), 7.58 (d, J = 5.6 Hz), 7.68 (d, 1H, J = 15.6 Hz), 8.37-8.40 (dt, 1H, J = 8.5,
2 Hz), 8.68 (s, 1H), 9.11 (s, 1H); 13C-NMR (DMSO-d6, 100 MHz) δ 116.11, 119.01, 123.55,
125.77, 131.01, 134.00, 135.86, 144.92, 149.86, 152.74, 160.90, 188.44; HMRS (esi) m/z calcd for
[M+H] 226.0868, found 226.0884.
4.1.2 General procedure for the synthesis of compounds 6-10
To a solution of the appropriate aldehyde (1 eq) in dichloromethane at room temperature (if the
aldehyde is not completely soluble it is possible to add some drops of DMF), pyridinium para-
toluenesulfonate was added as a catalyst, followed by a drop-wise addition of a solution of
dihydropyran (1 eq) in DCM. The mixture was stirred at room temperature until the starting
materials disappeared. The organic mixture was washed three times with water, dried over
magnesium sulphate and concentrated to give an orange to brown oil, which was used in the next
steps without further purification.
4.1.2.1 3-methoxy-4-((tetrahydro-2H-pyran-2-yl)oxy)benzaldehyde (6)
Compound 6 was obtained from 4-Hydroxy-3-methoxybenzaldehyde; yellow oil; HMRS (esi) m/z
calcd for [M+H] 237.1127, found 237.1143.
4.1.2.2 3-ethoxy-4-((tetrahydro-2H-pyran-2-yl)oxy)benzaldehyde (7)
Compound 7 was obtained from 4-Hydroxy-3-methoxybenzaldehyde; yellow oil; HMRS (esi) m/z
calcd for [M+H] 251.1283, found 251.1297.
4.1.2.3 3,4-bis((tetrahydro-2H-pyran-2-yl)oxy)benzaldehyde (8)
Compound 8 was obtained from 3,4-dihydroxybenzaldehyde; HMRS (esi) m/z calcd for [M+H]
307.1545, found 307.1531.
4.1.2.4 4-((tetrahydro-2H-pyran-2-yl)oxy)benzaldehyde (9)
Compound 9 was obtained from 3-dihydroxybenzaldehyde; yellow oil; HMRS (esi) m/z calcd for
[M+H] 207.1021, found 207.1045.
4.1.2.5 3-((tetrahydro-2H-pyran-2-yl)oxy)benzaldehyde (10)
20

ACCEPTED MANUSCRIPT
Compound 10 was obtained from 4-dihydroxybenzaldehyde; yellow oil; HMRS (esi) m/z calcd for
[M+H] 207.1021, found 207.1033.
4.1.3 General procedure for the synthesis of compounds 11-15
To a solution of the appropriate aldehydes 6-10 (1 eq) in NaOH aq 2.5 M, 3-acetylpyridine (1 eq)
was added dropwise and the color of the solution turned form from yellow to intense red; the
resulting mixture was stirred overnight at room temperature. The solvent was removed in vacuo and
the obtained residue was purified through flash chromatography.
4.1.3.1 (E)-3-(3-methoxy-4-((tetrahydro-2H-pyran-2-yl)oxy)phenyl)-1-(pyridin-3-yl)prop-2-en-1-
one (11)
Compound 11 was obtained from compound 6; eluting mixture dichloromethane / methanol (9:1);
1
yellow oil; 51% yield; H-NMR (CdCl3, 400 MHz) δ 1.61-1.68 (m, 2H), 1.77-1.88 (m, 2H), 1.91-
2.01 (m, 2H), 3.52-3.59 (m, 1H), 3.81-3.87 (m, 1H), 3.91 (s, 3H), 5.29-5.38 (m, 1H), 6.84 (d, 1H),
7.28-7.51 (d, 4H), 7.82-7.95 (d, 1H), 8.49 (d, 1H), 8.86 (m, 1H), 9.35 (d, 1H); HMRS (esi) m/z
calcd for [M+H] 340.1549, found 340.1557.
4.1.3.2 (E)-3-(3-ethoxy-4-((tetrahydro-2H-pyran-2-yl)oxy)phenyl)-1-(pyridin-3-yl)prop-2-en-1-one
(12)
Compound 12 was obtained from compound 7; eluting mixture dichloromethane / methanol (9:1);
1
yellow oil; 60% yield; H-NMR (CdCl3, 400 MHz) δ 1.42-1.56 (m, 4H), 1.61-1.71 (m, 2H), 1.72-
1.79 (m, 1H), 2.05-2.12 (m, 1H), 3.33-3.41 (m, 1H), 3.78-3.85 (m, 1H), 3.91 (s, 3H), 5.22-5.27 (m,
1H), 6.84 (d, 1H), 7.28-7.51 (d, 4H), 7.82-7.95 (d, 1H), 8.49 (d, 1H), 8.86 (m, 1H), 9.35 (d, 1H);
HMRS (esi) m/z calcd for [M+H] 354.1705, found 354.1698.
4.1.3.3 (E)-3-(3,4-bis((tetrahydro-2H-pyran-2-yl)oxy)phenyl)-1-(pyridin-3-yl)prop-2-en-1-one (13)
Compound 13 was obtained from compound 8; eluting mixture dichloromethane/methanol (9/1);
1
yellow oil; 45% yield; H-NMR (CdCl3, 400 MHz) δ 1.54-1.89 (m, 8H), 2.06.-2.47 (m, 4H), 3.77-
3.95 (m, 2H), 5.30-5.51 (m, 2H), 5.81-5.92 (m, 2H), 6.89 (d, 1H), 7.31-7.55 (d, 4H), 7.83-7.95 (d,
1H), 8.44 (d, 1H), 8.91 (m, 1H), 9.32 (d, 1H); HMRS (esi) m/z calcd for [M+H] 410.1967, found
410.1951.
4.1.3.4 (E)-1-(pyridin-3-yl)-3-(4-((tetrahydro-2H-pyran-2-yl)oxy)phenyl)prop-2-en-1-one (14)
Compound 14 was obtained from compound 9; eluting mixture dichloromethane / toluene /
1
methanol (8:1:1); yellow oil; 38 % yield; H-NMR (CdCl3, 400 MHz) δ 1.49-1.57 (m, 2H), 1.64-
1.69 (m, 2H), 1.77-1.99 (m, 2H), 3.79-3.84 (m, 1H), 3.91-3.95 (m, 1H), 5.89-5.94 (m, 1H), 6.92 (d,
1H), 7.34-7.58 (d, 4H), 7.84-91 (d, 2H), 8.32 (d, 1H), 8.82 (m, 1H), 9.23 (d, 1H); HMRS (esi) m/z
calcd for [M+H] 310.1443, found 310.1432.
21

ACCEPTED MANUSCRIPT
4.1.3.5 (E)-1-(pyridin-3-yl)-3-(3-((tetrahydro-2H-pyran-2-yl)oxy)phenyl)prop-2-en-1-one (15)
Compound 15 was obtained from compound 10; eluting mixture dichloromethane / methanol (9:1);
1
yellow oil; 42 % yield; H-NMR (CdCl3, 400 MHz) δ 1.53-1.62 (m, 2H), 1.61-1.73 (m, 2H), 1.81-
1.95 (m, 2H), 3.61-3.64 (m, 1H), 3.89-3.94 (m, 1H), 5.67-5.71 (m, 1H), 6.72 (d, 1H), 6.98 (d, 1H),
7.13- 7.33 (m, 4H), 7.60 (d, 1H), 7.82 (m, 1H), 8.35 (d, 1H), 8.81 (d, 1H); HMRS (esi) m/z calcd
for [M+H] 310.1443, found 310.1458.
4.1.4 Autoxidation experiments
Autoxidation experiments were performed in a two-channel oxygen-uptake apparatus based on a
Validyne DP 15 differential pressure transducer built in our laboratory and described
previously.[70-72] The antioxidant activity of the title compounds was evaluated by studying the
inhibition of the thermally initiated autoxidation of cumene (3.6 M) or styrene (4.3 M) in
chlorobenzene or THF (3.1 M) in aqueous phosphate buffer (pH 7.4). In a typical experiment, an
air-saturated mixture of the oxidizable substrate and the solvent, cumene or styrene/chlorobenzene
1:1 (v/v) containing AIBN (0.05 M) as initiator or THF/water 1:3 (v/v) containing the buffer (0.1M)
and AAPH (25mM) as initiator was equilibrated with an identical reference solution containing an
excess of 2,2,5,7,8-pentamethyl-6-chromanol. After equilibration, and when a constant O2
consumption was reached, 10-50 µL of a concentrated solution of the antioxidant was injected into
the sample flask and oxygen consumption in the sample was measured. From the slope of oxygen
consumption in the absence of antioxidant (-d[O2]/dt)0 = Rox0) and during the inhibited period (-
d[O2]/dt) = Rox), kinh values were obtained by fitting to Eq. (1), while the n coefficients were
determined from the length of the inhibited period (τ) by using Eq. (2), from the known rate of
radical production by AIBN or AAPH (initiation rate, Ri).[73]
Rox0 Rox nkinh
[
AH
]
0
-
=
Eq. (1)
Eq. (2)
Rox Rox0
2Kt Ri
R
iτ
n =
[
antioxidant
]
The 2kt values of cumene, styrene and THF at 303K are 4.2×107 x 107, 4.6×104 x 104 and 6.6×107 x
107 M-1s-1, respectively. The value of Ri was determined under each experimental condition by
using PMHC (2,2,5,7,8-pentamethyl-6-chromanol) or Trolox ((±)-6-Hydroxy-2,5,7,8-
tetramethylchromane-2-carboxylic acid) as a reference antioxidant: system = chlorobenzene /
cumene, Ri 4.3×10-9 M s-1; system = buffer pH 7.4/THF, Ri 7.3× x 10-9 M s-1.
22

ACCEPTED MANUSCRIPT
4.1.5 UV-vis Spectroscopy
UV-vis absorption spectra were recorded at room temperature by using a Jasco V550 double-beam
spectrometer with baseline correction. The solutions were placed in quartz absorption cuvettes with
a pathlength of 10 mm and a chamber volume of 3.5 mL. Stock solutions of compound 3 (1 mM)
and CuCl2 (1 mM) in methanol were freshly prepared and purged with nitrogen to prevent
oxidation. Under typical settings a fixed concentration of compound 3 (5 to 15 ×10-6 M) was added
of growing amounts of CuCl2 (up to 3 equivalents), oxygen was removed by purging with N2 and
the spectra (200-600 nm) were recorded vs neat methanol (solvent). At each concentration setting
the spectra of compound 3 and of CuCl2 at the same concentration in methanol were numerically
subtracted from that of the mixture to obtain the differential spectra. Molar extinction coefficients
for 3 in methanol were determined from calibration plots at 6 different concentrations (2-30 x 10-
6M) using Lambert-Beer's law.
4.2 Biology
4.2.1 Luminescence Assay for GSK-3β inhibition determination
Assays were performed in 50 mM HEPES, 1 mM EDTA, 1 mM EGTA, and 15 mM magnesium
acetate pH 7.5 assay buffer, using white 96-well plates. In a typical assay, 10 µL of test compound
(dissolved in DMSO at 1 mM concentration and diluted 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 GSM 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. After 10 min, luminescence in the entire visible range was
recorded using a Victor™ X3 Perkin Elmer multimode reader. The activity is proportional to the
difference between 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 SB-415826,[74] at total inhibition concentration,
respectively. The linear regression parameters were determined and the IC50 extrapolated
(GraphPad Prism 4.0, GraphPad Software Inc.).
Since the assay is based on the direct quantification of ATP cofactor by luciferase enzyme, for
compound 3 the activity was also evaluated in absence of GSK-3β and no interaction was detected
between compound 3 and luciferase up to 3 mM compound concentration.
23

ACCEPTED MANUSCRIPT
4.2.2 Determination of the inhibitory activity on Aβ42 self-aggregation
As reported in a previously published protocol,[41] a 1,1,1,3,3,3,-hexafluoro-2-propanol (HFIP)
pretreated Aβ42 sample (Bachem AG, Switzerland) was solubilized in a CH3CN/0.3 mM
Na2CO3/250 mM NaOH (48.4:48.4:3.2) mixture to obtain a 500 µM solution. Experiments were
performed by diluting the peptide to a final concentration of 50 µM with 10 mM phosphate buffer
(pH = 8.0) containing 10 mM NaCl, and incubating the diluted solution at 30°C. In these conditions
a reproducible aggregation kinetic was obtained.[41] For inhibition studies, Aβ42 (50 µM) was
incubated in 10 mM phosphate buffer (pH = 8.0) containing 10 mM NaCl, at 30°C (pH = 8.0) in the
absence and in the presence of compounds 1-5 (50 µM, Aβ/inhibitor = 1/1). Blanks containing the
inhibitors were also prepared and tested. Amyloid fibril formation was quantified by a ThT-
method.[69, 75]
Fluorescence intensities in the absence and in the presence of inhibitor were compared and the
percent inhibition due to the presence of the inhibitor was calculated. For the determination of the
IC50 value, increasing concentrations of 5, able to inhibit amyloid aggregation by 20-80%, were
evaluated.
To exclude any spectral interference between ThT and compound 5, fluorescence intensities
recorded before and after the addition of 5 to a pre-aggregated Aβ42 sample were compared.
4.2.3 Determination of the inhibitory activity on Tg2
Kinetic runs were recorded in triplicate on a BioTek Synergy H4 hybrid reader microplate reader in
absorbance mode at 405 nm and 25°C, in a buffer composed of 111 mM MOPS (pH 7.0), 3.33 mM
CaCl2, and 0.05 mM EDTA. All aqueous solutions were prepared using deionised water. Each
kinetic assay was performed using 900 µL buffer, 25 µL of a DMSO stock solution of substrate N-
Cbz-Glu(γ-p-nitrophenylester)Gly [76] (1 mg mL−1, 2.2 mM), 50 µL of a solution of recombinant
human TG2 (hTG2) at 0.05 U mL−1 (final concentration of 2.5 mU mL−1) and 0-25 µL of a DMSO
stock solution of inhibitor (contingent on solubility; the highest concentration tested ranged between
186 and 222 µM). The volume of DMSO was then adjusted so that it represented 5% of the final
volume. The reaction was initiated with the addition of hTG2 (expressed and purified according to a
literature protocol [77]), and the increase in absorbance was monitored over 10 minutes; initial[77]),
and the increase in absorbance was monitored over 10 minutes. Initial rates were measured over the
first minute of absorbance change, corresponding to the linear portion of the curve. IC50 values were
obtained as the negative x-intercept of a Dixon plot (reciprocal of initial rates vs. inhibitor
24

ACCEPTED MANUSCRIPT
concentration), from the average of two triplicate runs. The reported error was the standard error of
the mean.
4.2.4 Cell viability
For neuronal cytotoxicity studies, immature primary cortical neurons (E17) were isolated as
described[65] and plated at a density of 106 cells/ml in 96-well plates. The next day cells were
rinsed and then placed in medium containing 5 mM glutamate. Increasing concentrations of
compound 3 were added at the time of glutamate treatment. The next day, cell viability was
assessed by the MTT assay (Sigma).
4.2.5 PAMPA assay
To predict passive blood-brain penetration modification of the PAMPA has been used based on
reported protocol.[66, 67] The filter membrane of the donor plate was coated with PBL (Polar Brain
Lipid, Avanti, USA) in dodecane (4 µl of 20 mg/ml PBL in dodecane) and the acceptor well was
filled with 300 µl of PBS pH 7.4 buffer (VD). Tested compounds were dissolved first in DMSO and
that diluted with PBS pH 7.4 to reach the final concentration 100 µM in the donor well.
Concentration of DMSO did not exceed 0.5% (V/V) in the donor solution. 300 µL of the donor
solution was added to the donor wells (VA) and the donor filter plate was carefully put on the
acceptor plate so that coated membrane was “in touch” with both donor solution and acceptor
buffer. Test compound diffused from the donor well through the lipid membrane (Area = 0.28 cm2)
to the acceptor well. The concentration of the drug in both donor and the acceptor wells was
assessed after 3, 4, 5 and 6 h of incubation in quadruplicate using the UV plate reader Synergy HT
(Biotek, USA) at the maximum absorption wavelength of each compound. Concentration of the
compounds was calculated from the standard curve and expressed as permeability (Pe) value
according to Eq. 3:[78, 79]
ꢂ
ꢃ
ꢀ
×ꢀ
ꢀ
ꢀ
log
= log ꢀ × −
ꢁ1 − ꢂ
ꢄꢅ ℎꢀ ꢀ ꢀ ꢀ = ꢁꢆ
ꢄ (Eq. 3)
ꢀ ꢀ ꢀ ꢀ
ꢃ
ꢈ
×ꢀ ꢀ ꢀ ꢀ
ꢀ
ꢇ
ꢀ
×
ꢀ
ꢀ
Acknowledgements
This work was supported by the University of Bologna (RFO and FIRB FFBO123154 project
NANOX), the Italian Ministry for Education, Universities and Research (MIUR), The Alzheimer
Trento Onlus with the legato Baldrachi (to M.B and A.M.), the Natural Sciences and Engineering
25

ACCEPTED MANUSCRIPT
Research Council of Canada (to KYPA and JWK) and MH CZ - DRO (UHHK 00179906) for the
financial support. AM thanks Dr. Martina Battistini for excellent technical support.
Abbreviations
AChE, Acetylcholinesterase; AChEI, Acetylcholinesterase inhibitor; AD, Alzheimer’s disease
AIBN, Azobisisobutyronitrile; APP, amyloid precursor protein; Aβ, amyloid beta; BACE-1, β-
secretase 1; BBB, blood-brain barrier; BDE, bond dissociation enthalpy; CNS, central nervous
system; DIV1, day in vitro; ED, electron donating; EPT, electron-proton transfer; ET, electron
transfer; EW, electron withdrawing; FDA, food and drug administration; GSK-3β, glycogen
synthase kinase 3 β; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NFT,
neurofibrillary tangles; NMDA, N-methyl D-aspartate; PAMPA, parallel artificial membrane
permeation assay; PhCl, chlorobenzene; PT, proton transfer; ROS, reactive oxygen species; MTL,
multitarget ligand; SAR, structure-activity relationships; Tg2, transglutaminase 2; ThT, thioflavin T
Bibliography
[1] R. Morphy, Z. Rankovic, Designed multiple ligands. An emerging drug discovery paradigm, J Med Chem,
48 (2005) 6523-6543.
[2] M.L. Bolognesi, Polypharmacology in a single drug: multitarget drugs, Curr Med Chem, 20 (2013) 1639-
1645.
[3] A. Cavalli, M.L. Bolognesi, A. Minarini, M. Rosini, V. Tumiatti, M. Recanatini, C. Melchiorre, Multi-target-
directed ligands to combat neurodegenerative diseases, J Med Chem, 51 (2008) 347-372.
[4] E. Karran, M. Mercken, B. De Strooper, The amyloid cascade hypothesis for Alzheimer's disease: an
appraisal for the development of therapeutics, Nat Rev Drug Discov, 10 (2011) 698-712.
[5] R.B. Maccioni, G. Farías, I. Morales, L. Navarrete, The revitalized tau hypothesis on Alzheimer's disease,
Arch Med Res, 41 (2010) 226-231.
[6] J.J. Palop, J. Chin, L. Mucke, A network dysfunction perspective on neurodegenerative diseases, Nature,
443 (2006) 768-773.
[7] A. Wimo, and Prince, M., World Alzheimer Report 2010: The Global EconomicImpact of Dementia,
London: Alzheimer’s Disease International, 2010, pp. 1–56.
[8] J.L. Cummings, T. Morstorf, K. Zhong, Alzheimer's disease drug-development pipeline: few candidates,
frequent failures, Alzheimers Res Ther, 6 (2014) 37.
[9] S.O. Bachurin, E.V. Bovina, A.A. Ustyugov, Drugs in Clinical Trials for Alzheimer's Disease: The Major
Trends, Med Res Rev, (2017).
[10] L. Ismaili, B. Refouvelet, M. Benchekroun, S. Brogi, M. Brindisi, S. Gemma, G. Campiani, S. Filipic, D.
Agbaba, G. Esteban, M. Unzeta, K. Nikolic, S. Butini, J. Marco-Contelles, Multitarget compounds bearing
tacrine- and donepezil-like structural and functional motifs for the potential treatment of Alzheimer's
disease, Prog Neurobiol, (2016).
[11] V. Tumiatti, A. Minarini, M.L. Bolognesi, A. Milelli, M. Rosini, C. Melchiorre, Tacrine derivatives and
Alzheimer's disease, Curr Med Chem, 17 (2010) 1825-1838.
[12] A. Minarini, A. Milelli, E. Simoni, M. Rosini, M.L. Bolognesi, C. Marchetti, V. Tumiatti, Multifunctional
tacrine derivatives in Alzheimer's disease, Curr Top Med Chem, 13 (2013) 1771-1786.
[13] F. Prati, A. De Simone, P. Bisignano, A. Armirotti, M. Summa, D. Pizzirani, R. Scarpelli, D.I. Perez, V.
Andrisano, A. Perez-Castillo, B. Monti, F. Massenzio, L. Polito, M. Racchi, A.D. Favia, G. Bottegoni, A.
Martinez, M.L. Bolognesi, A. Cavalli, Multitarget drug discovery for Alzheimer's disease: triazinones as
BACE-1 and GSK-3β inhibitors, Angew Chem Int Ed Engl, 54 (2015) 1578-1582.
26

ACCEPTED MANUSCRIPT
[14] F. Prati, A. De Simone, A. Armirotti, M. Summa, D. Pizzirani, R. Scarpelli, S.M. Bertozzi, D.I. Perez, V.
Andrisano, A. Perez-Castillo, B. Monti, F. Massenzio, L. Polito, M. Racchi, P. Sabatino, G. Bottegoni, A.
Martinez, A. Cavalli, M.L. Bolognesi, 3,4-Dihydro-1,3,5-triazin-2(1H)-ones as the First Dual BACE-1/GSK-3β
Fragment Hits against Alzheimer's Disease, ACS Chem Neurosci, 6 (2015) 1665-1682.
[15] R. Yan, R. Vassar, Targeting the β secretase BACE1 for Alzheimer's disease therapy, Lancet Neurol, 13
(2014) 319-329.
[16] C. Hooper, R. Killick, S. Lovestone, The GSK3 hypothesis of Alzheimer's disease, J Neurochem, 104
(2008) 1433-1439.
[17] M. Llorens-Martín, J. Jurado, F. Hernández, J. Avila, GSK-3β, a pivotal kinase in Alzheimer disease, Front
Mol Neurosci, 7 (2014) 46.
[18] M.E. Kennedy, A.W. Stamford, X. Chen, K. Cox, J.N. Cumming, M.F. Dockendorf, M. Egan, L. Ereshefsky,
R.A. Hodgson, L.A. Hyde, S. Jhee, H.J. Kleijn, R. Kuvelkar, W. Li, B.A. Mattson, H. Mei, J. Palcza, J.D. Scott, M.
Tanen, M.D. Troyer, J.L. Tseng, J.A. Stone, E.M. Parker, M.S. Forman, The BACE1 inhibitor verubecestat (MK-
8931) reduces CNS β-amyloid in animal models and in Alzheimer's disease patients, Sci Transl Med, 8 (2016)
363ra150.
[19] H. Ma, S. Lesné, L. Kotilinek, J.V. Steidl-Nichols, M. Sherman, L. Younkin, S. Younkin, C. Forster, N.
Sergeant, A. Delacourte, R. Vassar, M. Citron, P. Kofuji, L.M. Boland, K.H. Ashe, Involvement of beta-site
APP cleaving enzyme 1 (BACE1) in amyloid precursor protein-mediated enhancement of memory and
activity-dependent synaptic plasticity, Proc Natl Acad Sci U S A, 104 (2007) 8167-8172.
[20] S. Barão, D. Moechars, S.F. Lichtenthaler, B. De Strooper, BACE1 Physiological Functions May Limit Its
Use as Therapeutic Target for Alzheimer's Disease, Trends Neurosci, 39 (2016) 158-169.
[21] T. Mohamed, A. Shakeri, P.P. Rao, Amyloid cascade in Alzheimer's disease: Recent advances in
medicinal chemistry, Eur J Med Chem, 113 (2016) 258-272.
[22] J. Sevigny, P. Chiao, T. Bussière, P.H. Weinreb, L. Williams, M. Maier, R. Dunstan, S. Salloway, T. Chen,
Y. Ling, J. O'Gorman, F. Qian, M. Arastu, M. Li, S. Chollate, M.S. Brennan, O. Quintero-Monzon, R.H.
Scannevin, H.M. Arnold, T. Engber, K. Rhodes, J. Ferrero, Y. Hang, A. Mikulskis, J. Grimm, C. Hock, R.M.
Nitsch, A. Sandrock, The antibody aducanumab reduces Aβ plaques in Alzheimer's disease, Nature, 537
(2016) 50-56.
[23] D.J. Selkoe, J. Hardy, The amyloid hypothesis of Alzheimer's disease at 25 years, EMBO Mol Med, 8
(2016) 595-608.
[24] M. Rosini, E. Simoni, A. Milelli, A. Minarini, C. Melchiorre, Oxidative stress in Alzheimer's disease: are
we connecting the dots?, J Med Chem, 57 (2014) 2821-2831.
[25] G. Perry, A.D. Cash, M.A. Smith, Alzheimer Disease and Oxidative Stress, J Biomed Biotechnol, 2 (2002)
120-123.
[26] Y. Zhao, B. Zhao, Oxidative stress and the pathogenesis of Alzheimer's disease, Oxid Med Cell Longev,
2013 (2013) 316523.
[27] S.M. Alavi Naini, N. Soussi-Yanicostas, Tau Hyperphosphorylation and Oxidative Stress, a Critical Vicious
Circle in Neurodegenerative Tauopathies?, Oxid Med Cell Longev, 2015 (2015) 151979.
[28] B. Su, X. Wang, H.G. Lee, M. Tabaton, G. Perry, M.A. Smith, X. Zhu, Chronic oxidative stress causes
increased tau phosphorylation in M17 neuroblastoma cells, Neurosci Lett, 468 (2010) 267-271.
[29] M.A. Lovell, S. Xiong, C. Xie, P. Davies, W.R. Markesbery, Induction of hyperphosphorylated tau in
primary rat cortical neuron cultures mediated by oxidative stress and glycogen synthase kinase-3, J
Alzheimers Dis, 6 (2004) 659-671; discussion 673-681.
[30] C.X. Peng, J. Hu, D. Liu, X.P. Hong, Y.Y. Wu, L.Q. Zhu, J.Z. Wang, Disease-modified glycogen synthase
kinase-3β intervention by melatonin arrests the pathology and memory deficits in an Alzheimer's animal
model, Neurobiol Aging, 34 (2013) 1555-1563.
[31] G. Scapagnini, D.A. Butterfield, C. Colombrita, R. Sultana, A. Pascale, V. Calabrese, Ethyl ferulate, a
lipophilic polyphenol, induces HO-1 and protects rat neurons against oxidative stress, Antioxid Redox
Signal, 6 (2004) 811-818.
[32] B.A. Cussac M, Chimica Therapeutica, 1967, pp. 101.
[33] JpnKokai Tokkyo Koho (1992), JP 04154720 A 19920527.
27

ACCEPTED MANUSCRIPT
[34] C. Zhuang, W. Zhang, C. Sheng, C. Xing, Z. Miao, Chalcone: A Privileged Structure in Medicinal
Chemistry, Chem Rev, (2017).
[35] R.M. Di Martino, A. De Simone, V. Andrisano, P. Bisignano, A. Bisi, S. Gobbi, A. Rampa, R. Fato, C.
Bergamini, D.I. Perez, A. Martinez, G. Bottegoni, A. Cavalli, F. Belluti, Versatility of the Curcumin Scaffold:
Discovery of Potent and Balanced Dual BACE-1 and GSK-3β Inhibitors, J Med Chem, 59 (2016) 531-544.
[36] A.A. Reinke, J.E. Gestwicki, Structure-activity relationships of amyloid beta-aggregation inhibitors
based on curcumin: influence of linker length and flexibility, Chem Biol Drug Des, 70 (2007) 206-215.
[37] C. Pardin, J.N. Pelletier, W.D. Lubell, J.W. Keillor, Cinnamoyl inhibitors of tissue transglutaminase, J Org
Chem, 73 (2008) 5766-5775.
[38] M. Lesort, J. Tucholski, M.L. Miller, G.V. Johnson, Tissue transglutaminase: a possible role in
neurodegenerative diseases, Prog Neurobiol, 61 (2000) 439-463.
[39] A. Baki, A. Bielik, L. Molnár, G. Szendrei, G.M. Keserü, A high throughput luminescent assay for
glycogen synthase kinase-3beta inhibitors, Assay Drug Dev Technol, 5 (2007) 75-83.
[40] L. Avrahami, A. Licht-Murava, M. Eisenstein, H. Eldar-Finkelman, GSK-3 inhibition: achieving moderate
efficacy with high selectivity, Biochim Biophys Acta, 1834 (2013) 1410-1414.
[41] M. Bartolini, C. Bertucci, M.L. Bolognesi, A. Cavalli, C. Melchiorre, V. Andrisano, Insight into the kinetic
of amyloid beta (1-42) peptide self-aggregation: elucidation of inhibitors' mechanism of action,
Chembiochem, 8 (2007) 2152-2161.
[42] Q.I. Churches, J. Caine, K. Cavanagh, V.C. Epa, L. Waddington, C.E. Tranberg, A.G. Meyer, J.N. Varghese,
V. Streltsov, P.J. Duggan, Naturally occurring polyphenolic inhibitors of amyloid beta aggregation, Bioorg
Med Chem Lett, 24 (2014) 3108-3112.
[43] X. Chao, X. He, Y. Yang, X. Zhou, M. Jin, S. Liu, Z. Cheng, P. Liu, Y. Wang, J. Yu, Y. Tan, Y. Huang, J. Qin, S.
Rapposelli, R. Pi, Design, synthesis and pharmacological evaluation of novel tacrine-caffeic acid hybrids as
multi-targeted compounds against Alzheimer's disease, Bioorg Med Chem Lett, 22 (2012) 6498-6502.
[44] A. Minarini, A. Milelli, V. Tumiatti, M. Rosini, E. Simoni, M.L. Bolognesi, V. Andrisano, M. Bartolini, E.
Motori, C. Angeloni, S. Hrelia, Cystamine-tacrine dimer: a new multi-target-directed ligand as potential
therapeutic agent for Alzheimer's disease treatment, Neuropharmacology, 62 (2012) 997-1003.
[45] E. Simoni, M.M. Serafini, M. Bartolini, R. Caporaso, A. Pinto, D. Necchi, J. Fiori, V. Andrisano, A.
Minarini, C. Lanni, M. Rosini, Nature-Inspired Multifunctional Ligands: Focusing on Amyloid-Based
Molecular Mechanisms of Alzheimer's Disease, ChemMedChem, 11 (2016) 1309-1317.
[46] S. Ayton, P. Lei, A.I. Bush, Biometals and their therapeutic implications in Alzheimer's disease,
Neurotherapeutics, 12 (2015) 109-120.
[47] J.S. Cristóvão, R. Santos, C.M. Gomes, Metals and Neuronal Metal Binding Proteins Implicated in
Alzheimer's Disease, Oxid Med Cell Longev, 2016 (2016) 9812178.
[48] X. Huang, M.P. Cuajungco, C.S. Atwood, M.A. Hartshorn, J.D. Tyndall, G.R. Hanson, K.C. Stokes, M.
Leopold, G. Multhaup, L.E. Goldstein, R.C. Scarpa, A.J. Saunders, J. Lim, R.D. Moir, C. Glabe, E.F. Bowden,
C.L. Masters, D.P. Fairlie, R.E. Tanzi, A.I. Bush, Cu(II) potentiation of alzheimer abeta neurotoxicity.
Correlation with cell-free hydrogen peroxide production and metal reduction, J Biol Chem, 274 (1999)
37111-37116.
[49] C.J. Sarell, S.R. Wilkinson, J.H. Viles, Substoichiometric levels of Cu2+ ions accelerate the kinetics of
fiber formation and promote cell toxicity of amyloid-{beta} from Alzheimer disease, J Biol Chem, 285 (2010)
41533-41540.
[50] v.d.B.M.G. Constant, Determination of the complexing capacity and conditional stability constants of
complexes of copper(II) with natural organic ligands in seawater by cathodic stripping voltammetry of
copper-catechol complex ions, Marine Chemistry, 1984, pp. 1-18.
[51] R. Amorati, L. Valgimigli, Advantages and limitations of common testing methods for antioxidants, Free
Radic Res, 49 (2015) 633-649.
[52] R. Matera, S. Gabbanini, S. Berretti, R. Amorati, G.R. De Nicola, R. Iori, L. Valgimigli, Acylated
anthocyanins from sprouts of Raphanus sativus cv. Sango: isolation, structure elucidation and antioxidant
activity, Food Chem, 166 (2015) 397-406.
[53] R. Amorati, L. Valgimigli, Modulation of the antioxidant activity of phenols by non-covalent
interactions, Org Biomol Chem, 10 (2012) 4147-4158.
28

ACCEPTED MANUSCRIPT
[54] R. Amorati, S. Menichetti, E. Mileo, G.F. Pedulli, C. Viglianisi, Hydrogen-atom transfer reactions from
ortho-alkoxy-substituted phenols: an experimental approach, Chemistry, 15 (2009) 4402-4410.
[55] L. Valgimigli, D.A. Pratt, Maximizing the reactivity of phenolic and aminic radical-trapping antioxidants:
just add nitrogen!, Acc Chem Res, 48 (2015) 966-975.
[56] R. Amorati, G.F. Pedulli, L. Cabrini, L. Zambonin, L. Landi, Solvent and pH effects on the antioxidant
activity of caffeic and other phenolic acids, J Agric Food Chem, 54 (2006) 2932-2937.
[57] R. Amorati, L. Valgimigli, L. Panzella, A. Napolitano, M. d'Ischia, 5-S-lipoylhydroxytyrosol, a
multidefense antioxidant featuring
a
solvent-tunable peroxyl radical-scavenging 3-thio-1,2-
dihydroxybenzene motif, J Org Chem, 78 (2013) 9857-9864.
[58] R. Amorati, A. Baschieri, G. Morroni, R. Gambino, L. Valgimigli, Peroxyl Radical Reactions in Water
Solution: A Gym for Proton-Coupled Electron-Transfer Theories, Chemistry, 22 (2016) 7924-7934.
[59] W. Zhang, B.R. Johnson, D.E. Suri, J. Martinez, T.D. Bjornsson, Immunohistochemical demonstration of
tissue transglutaminase in amyloid plaques, Acta Neuropathol, 96 (1998) 395-400.
[60] B.A. Citron, Z. Suo, K. SantaCruz, P.J. Davies, F. Qin, B.W. Festoff, Protein crosslinking, tissue
transglutaminase, alternative splicing and neurodegeneration, Neurochem Int, 40 (2002) 69-78.
[61] S.M. Dudek, G.V. Johnson, Transglutaminase facilitates the formation of polymers of the beta-amyloid
peptide, Brain Res, 651 (1994) 129-133.
[62] M.L. Miller, G.V. Johnson, Transglutaminase cross-linking of the tau protein, J Neurochem, 65 (1995)
1760-1770.
[63] J.J. Wakshlag, M.A. Antonyak, J.E. Boehm, K. Boehm, R.A. Cerione, Effects of tissue transglutaminase
on beta -amyloid1-42-induced apoptosis, Protein J, 25 (2006) 83-94.
[64] R.R. Ratan, T.H. Murphy, J.M. Baraban, Oxidative stress induces apoptosis in embryonic cortical
neurons, J Neurochem, 62 (1994) 376-379.
[65] M. Basso, J. Berlin, L. Xia, S.F. Sleiman, B. Ko, R. Haskew-Layton, E. Kim, M.A. Antonyak, R.A. Cerione,
S.E. Iismaa, D. Willis, S. Cho, R.R. Ratan, Transglutaminase inhibition protects against oxidative stress-
induced neuronal death downstream of pathological ERK activation, J Neurosci, 32 (2012) 6561-6569.
[66] L. Di, E.H. Kerns, K. Fan, O.J. McConnell, G.T. Carter, High throughput artificial membrane permeability
assay for blood-brain barrier, Eur J Med Chem, 38 (2003) 223-232.
[67] E. Nepovimova, J. Korabecny, R. Dolezal, T.D. Nguyen, D. Jun, O. Soukup, M. Pasdiorova, P. Jost, L.
Muckova,
pyridinealdoxime hybrid
D. Malinak,
L.
as
Gorecki,
novel
K. Musilek,
K. Kuca,
A
7-methoxytacrine-4-
a
prophylactic agent with reactivation properties in
organophosphate intoxication, Toxicology Research, 2016, pp. 1012-1016.
[68] L.F. Lemes, G. de Andrade Ramos, A.S. de Oliveira, F.M. da Silva, G. de Castro Couto, M. da Silva Boni,
M.J. Guimarães, I.N. Souza, M. Bartolini, V. Andrisano, P.C. do Nascimento Nogueira, E.R. Silveira, G.D.
Brand, O. Soukup, J. Korábečný, N.C. Romeiro, N.G. Castro, M.L. Bolognesi, L.A. Romeiro, Cardanol-derived
AChE inhibitors: Towards the development of dual binding derivatives for Alzheimer's disease, Eur J Med
Chem, 108 (2016) 687-700.
[69] H. LeVine, Thioflavine T interaction with synthetic Alzheimer's disease beta-amyloid peptides:
detection of amyloid aggregation in solution, Protein Sci, 2 (1993) 404-410.
[70] R. Amorati, L. Valgimigli, P. Dinér, K. Bakhtiari, M. Saeedi, L. Engman, Multi-faceted reactivity of
alkyltellurophenols towards peroxyl radicals: catalytic antioxidant versus thiol-depletion effect, Chemistry,
19 (2013) 7510-7522.
[71] R. Amorati, J. Zotova, A. Baschieri, L. Valgimigli, Antioxidant Activity of Magnolol and Honokiol: Kinetic
and Mechanistic Investigations of Their Reaction with Peroxyl Radicals, J Org Chem, 80 (2015) 10651-10659.
[72] L. Valgimigli, D. Bartolomei, R. Amorati, E. Haidasz, J.J. Hanthorn, S.J. Nara, J. Brinkhorst, D.A. Pratt, 3-
Pyridinols and 5-pyrimidinols: Tailor-made for use in synergistic radical-trapping co-antioxidant systems,
Beilstein J Org Chem, 9 (2013) 2781-2792.
[73] R. Amorati, G.F. Pedulli, L. Valgimigli, Kinetic and thermodynamic aspects of the chain-breaking
antioxidant activity of ascorbic acid derivatives in non-aqueous media, Org Biomol Chem, 9 (2011) 3792-
3800.
[74] M.P. Coghlan, A.A. Culbert, D.A. Cross, S.L. Corcoran, J.W. Yates, N.J. Pearce, O.L. Rausch, G.J. Murphy,
P.S. Carter, L. Roxbee Cox, D. Mills, M.J. Brown, D. Haigh, R.W. Ward, D.G. Smith, K.J. Murray, A.D. Reith,
29