Merging memantine and ferulic acid to probe connections between NMDA receptors, oxidative stress and amyloid-β peptide in Alzheimer's disease

Article abstract of DOI:10.1016/j.ejmech.2019.07.011

N-methyl-D-aspartate receptors (NMDAR) are critically involved in the pathogenesis of Alzheimer's disease (AD). Acting as an open-channel blocker, the anti-AD drug memantine preferentially targets NMDAR overactivation, which has been proposed to trigger neurotoxic events mediated by amyloid β peptide (Aβ) and oxidative stress. In this study, we applied a multifunctional approach by conjugating memantine to ferulic acid, which is known to protect the brain from Aβ neurotoxicity and neuronal death caused by ROS. The most interesting compound (7) behaved, like memantine, as a voltage-dependent antagonist of NMDAR (IC₅₀ = 6.9 μM). In addition, at 10 μM concentration, 7 exerted antioxidant properties both directly and indirectly through the activation of the Nrf-2 pathway in SH-SY5Y cells. At the same concentration, differently from the parent compounds memantine and ferulic acid alone, it was able to modulate Aβ production, as revealed by the observed increase of the non-amyloidogenic sAPPα in H4-SW cells. These findings suggest that compound 7 may represent a promising tool for investigating NMDAR-mediated neurotoxic events involving Aβ burden and oxidative damage.

Full text of DOI:10.1016/j.ejmech.2019.07.011

Accepted Manuscript
Merging memantine and ferulic acid to probe connections between NMDA receptors,
oxidative stress and amyloid-β peptide in Alzheimer's disease
Michela Rosini, Elena Simoni, Roberta Caporaso, Filippo Basagni, Michele
Catanzaro, Izuddin F. Abu, Francesca Fagiani, Federica Fusco, Sara Masuzzo, Diego
Albani, Cristina Lanni, Ian R. Mellor, Anna Minarini
PII:
S0223-5234(19)30632-4
DOI:
https://doi.org/10.1016/j.ejmech.2019.07.011
EJMECH 11508
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 10 June 2019
Revised Date: 2 July 2019
Accepted Date: 3 July 2019
Please cite this article as: M. Rosini, E. Simoni, R. Caporaso, F. Basagni, M. Catanzaro, I.F. Abu, F.
Fagiani, F. Fusco, S. Masuzzo, D. Albani, C. Lanni, I.R. Mellor, A. Minarini, Merging memantine and
ferulic acid to probe connections between NMDA receptors, oxidative stress and amyloid-β peptide
in Alzheimer's disease, European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/
j.ejmech.2019.07.011.
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

ACCEPTED MANUSCRIPT
Merging memantine and ferulic acid to probe connections between NMDA
receptors, oxidative stress and amyloid-β peptide in Alzheimer’s disease
Michela Rosini1,*, Elena Simoni1, Roberta Caporaso1, Filippo Basagni1, Michele Catanzaro2, Izuddin F. Abu3,4,
Francesca Fagiani2,5, Federica Fusco6, Sara Masuzzo6, Diego Albani6, Cristina Lanni2, Ian R. Mellor3 and Anna
Minarini1,*
1 Department of Pharmacy and Biotechnology, Alma Mater Studiorum – University of Bologna, Via
Belmeloro 6, 40126 Bologna (IT)
2 Department of Drug Sciences (Pharmacology Section), University of Pavia, V.le Taramelli 14, 27100 Pavia
(IT)
3 School of Life Sciences, University of Nottingham, University Park, Nottingham, NG7 2RD (UK)
4 Universiti Kuala Lumpur, Institute of Medical Science Technology, A1-1, Jalan TKS1, Taman Kajang Sentral,
43000 Kajang, Selangor, Malaysia
5 Scuola Universitaria Superiore IUSS Pavia, P.zza Vittoria, 15, 27100 Pavia (IT)
6 Department of Neuroscience, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Via Mario Negri 2,
20156 Milan (IT)
Corresponding Authors
* Michela Rosini: michela.rosini@unibo.it, Phone: +39 051 2099722; Anna Minarini:
anna.minarini@unibo.it, Phone: +39 051 2099709.
1

ACCEPTED MANUSCRIPT
ABSTRACT
N-methyl-D-aspartate receptors (NMDAR) are critically involved in the pathogenesis of Alzheimer’s disease
(AD). Acting as an open-channel blocker, the anti-AD drug memantine preferentially targets NMDAR
overactivation, which has been proposed to trigger neurotoxic events mediated by amyloid β peptide (Aβ)
and oxidative stress. In this study, we applied a multifunctional approach by conjugating memantine to
ferulic acid, which is known to protect the brain from Aβ neurotoxicity and neuronal death caused by ROS.
The most interesting compound (7) behaved, like memantine, as a voltage-dependent antagonist of
NMDAR (IC50= 6.9 µM). In addition, at 10 µM concentration, 7 exerted antioxidant properties both directly
and indirectly through the activation of the Nrf-2 pathway in SH-SY5Y cells. At the same concentration,
differently from the parent compounds memantine and ferulic acid alone, it was able to modulate
Aβ production, as revealed by the observed increase of the non-amyloidogenic sAPPα in H4-sw cells. These
findings suggest that compound 7 may represent a promising tool for investigating NMDAR-mediated
neurotoxic events involving Aβ burden and oxidative damage.
Keywords
Memantine, ferulic acid, NMDA receptors, oxidative stress, amyloid-β peptide
Abbreviations
AD: Alzheimer’s disease; ANOVA: analysis of variance; APP: amyloid precursor protein; Aβ: amyloid β
peptide; DCFH-DA: dichlorodihydrofluorescein diacetate; eNMDAR: extrasynaptic NMDAR; FA: ferulic acid;
fl-APP: full-length APP; H4-sw: APP-Swedish mutant-expressing H4 cell lines; HO-1: heme oxygenase-1;
MEM: memantine; MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; NMDAR: N-methyl-
D-aspartate receptors; Nrf2: nuclear factor (erythroid-derived 2)-like 2; Ppm: parts per millions; sAPPα:
soluble amyloid precursor protein α; SDS: sodium dodecyl sulfate; sNMDAR: synaptic NMDAR; TMS:
tetramethylsilane.
2

ACCEPTED MANUSCRIPT
INTRODUCTION
Synaptic loss is a major feature in Alzheimer’s disease (AD). This chronic neurodegenerative condition,
which is currently afflicting about 47 million people worldwide, slowly destroys neurons leading to
progressive cognitive disabilities [1]. How synapses are affected in the disease process remains unclear. The
mechanistic understanding of synaptic damage represents a challenging goal, and may offer new
possibilities for the prevention and cure of the disease. N-methyl-D-aspartate receptors (NMDAR) are
ionotropic glutamate receptors known to play an important role for synaptic plasticity in the healthy brain
[2]. They are primarily involved in neuronal excitatory synaptic transmission that underlies learning and
memory but also in excitotoxic damage occurring during acute brain injuries and chronic
neurodegenerative conditions. Targeting NMDAR therapeutically is therefore complicated by the
dichotomous nature of their downstream signaling. It is the common view that these opposite effects
depend on receptor localization: activation of synaptic NMDAR (sNMDAR) may contribute to cell survival
and plasticity, while activation of extrasynaptic NMDAR (eNMDAR) may preferentially signal to neuronal
death [3,4]. In particular, overactivation of eNMDAR has been associated with glutamate-mediated
oxidative damage potentially leading to aberrant, misfolded proteins [5]. The amyloid β peptide (Aβ) is a
pathogenic feature of AD development. Produced by the sequential cleavage of the amyloid precursor
protein (APP) by β- and γ-secretases, as an alternative to the non-amyloidogenic cleavage performed by α-
secretase, Aβ monomers aggregate into soluble oligomeric forms, which are believed to be mainly
responsible for amyloid-driven synaptotoxicity [6]. A toxic positive feedback is established between Aβ
production and eNMDAR overactivation, which involves cytoplasmic Ca2+ upregulation and aberrant redox–
mediated reactions [7].
Memantine is an anti-AD drug currently in use for the treatment of moderate-to-severe forms of the
disease. It is an uncompetitive/fast off-rate NMDAR antagonist. By acting as an open-channel blocker, it
preferentially enters the channel’s pore in conditions of excessive and prolonged glutamate exposure [8,9].
Its favorable kinetics has been proposed to selectively direct memantine’s efficacy toward
extrasynaptic/tonically-activated NMDAR over synaptic/phasically-activated NMDAR [10], accounting for
3

ACCEPTED MANUSCRIPT
the clinical tolerability of the drug. Further, this peculiar profile seems to play a crucial role in determining
memantine’s ability to alleviate Aβ-induced synaptic dysfunction and to rescue both neuronal oxidative
stress and the transient memory impairment caused by Aβ oligomers [11]. Unfortunately, however, like
other available anti-AD drugs, memantine offers only a symptomatic relief to patients and is not able to
halt the disease progression.
Based on these premises, we sought to combine in a single molecule memantine, which specifically
modulates NMDAR-mediated excitotoxicity, responsible for ROS- and Aβ-mediated neurotoxic events, with
the antioxidant ferulic acid (FA), whose well-established biological properties include the ability to protect
the brain from Aβ neurotoxicity and neuronal death caused by ROS [12]. Following this rationale, we
designed and synthesized memantine-FA conjugates following the two routes shown in Figure 1.
Figure 1. Drug design of compounds 1-7.
It is well known that memantine’s primary amine plays a crucial role in receptor binding [13]. Thus, to
preserve this moiety, we functionalized the adamantane nucleus of memantine with a carboxylic function,
4

ACCEPTED MANUSCRIPT
which acted as the reactive point for FA conjugation, affording compounds 1-3. Further, in compounds 4-7,
we explored the possibility to introduce FA appendages on the nitrogen atom of memantine, whose
conversion to a secondary amine has previously emerged as a feasible strategy to gain memantine-based
NMDAR antagonists [14,15]. Synthesized compounds were first tested against NMDAR. Based on their
NMDAR blocking properties, compounds were selected to study their direct and indirect antioxidant
efficacy, as well as the ability to modulate the amyloidogenic pathway.
RESULTS AND DISCUSSION
Chemistry. Memantine-FA hybrids 1-7 were prepared following two different synthetic routes (Schemes 1
and 2), depending on the way the two pharmacophores were connected. For the synthesis of compounds
1-3, the appropriate mono Boc-protected diamine (8-10) [16,17] was condensed with FA to give
intermediates 11-13. Cleavage of the protecting group in acidic conditions led to compounds 14-16.
Conjugation of 14-16 with 17 hydrochloride, which was obtained following a Ritter-type protocol as
previously reported by Wanka et al. [18], afforded final compounds 1-3 (Scheme 1). To gain compounds 4-
7, memantine hydrochloride (MEM) was alkylated with the appropriate tosyl-activated alcohol (18-21)
under basic conditions to give intermediates 22-25 which, after carbamate deprotection (26-29), were
coupled with FA in the presence of EDC and HOBt (Scheme 2).
5

ACCEPTED MANUSCRIPT
Scheme 1. Reagents and conditions: (a) EDC, HOBt, DMF, Et3N, N2, 12 h, 0°C-rt; (b) HCl 4 M in dioxane,
CH2Cl2, 90’, 0°C-rt; (c) EDC, HOBt, DMF, N2, 36 h, 0°C-rt.
Scheme 2. Reagents and conditions: (a) K2CO3, KI, DMF, 140°C, 1h, MW; (b) HCl 4 M in dioxane, CH2Cl2, 20’,
0°C-rt; (c) EDC, HOBt, Et3N, DMF, N2, 12 h, 0°C-rt.
NMDAR blocking activity. All the compounds were initially investigated to assess their effect at NMDAR. In
particular, the antagonism of responses to NMDA and glycine were measured by voltage-clamp recordings
on GluN1-1a/GluN2A NMDAR expressed in Xenopus laevis oocytes at -60 mV, with compounds 1-7 applied
in tenfold increments in the range 0.01 to 100 μM. Memantine was used as the reference compound.
Compounds 1-3, which retain the primary amine function of memantine, demonstrated very low or no
potency to block NMDAR (Figure 2A). Conversely, employing memantine’s amine for connecting FA
appendages resulted in significant blocking of NMDA/glycine responses (Figure 2A).
6

ACCEPTED MANUSCRIPT
Figure 2. A) Concentration-inhibition curves for compounds 1-7 in comparison to memantine (Mem). Data
are mean % of control response to 100 μM NMDA (+10 μM glycine) ± SEM (n = 5-7 separate oocytes). The
curves are fits to Equation 1 and IC50 values are given in Table 1. B) Voltage dependence of inhibition by
compounds 5-7 (30 μM, 20 μM and 10 μM respectively) in comparison to memantine (Mem; 3 μM). Data
are plotted as mean % control response to 100 μM NMDA (+10 μM glycine) ± SEM against the holding
potential (Vh) (n = 5-6 separate oocytes). The curves are fits of Equation 2 and δ values are given in Table 1.
Compounds 5-7 presented a micromolar profile, with IC50 values ranging from 6.9 to 23.9 µM, while the
shorter compound 4 had an IC50 greater than 100 μM, thus suggesting its inefficacy (Table 1). Blocking
properties toward NMDAR were influenced by the chain length separating the pharmacophoric functions,
with compound 7, carrying a hexamethylene spacer, emerging as the most efficacious. Compounds with
considerable blocking properties (5-7) were assessed for voltage dependency. The compounds were diluted
7

ACCEPTED MANUSCRIPT
to their approximate IC50 concentrations and block of NMDA/glycine responses mediated by GluN1-
1a/GluN2A was measured at four different holding potentials (-40, -60, -80 and -100 mV).
Table 1. IC50 and δ values for compounds 1-7 and memantine
Compound
IC50 [95% CI] μM (n)
>>100 (7)
δ ± SE (n)
nd
1
2
>>100 (6)
nd
3
>>100 (6)
nd
4
>100 (5)
nd
5
23.9 [13.0-49.4] (5)
14.1 [8.7-22.6] (5)
6.9 [3.0-19.2] (5)
2.3 [1.7-3.0] (6)
0.46 ± 0.07 (6)
0.43 ± 0.12 (6)
0.51 ± 0.17 (6)
0.39 ± 0.08 (5)
6
7
memantine
nd = not determined (because inhibition was too weak).
Compounds presented a voltage-dependent behavior, acting, like memantine, as open channel blockers of
the receptor. Data were fitted with the Woodhull equation to determine their δ values, and thus estimate
the position of the binding site within the membrane electric field [19,20]. The results of this study showed
that three of the new molecules yielded δ values comparable to memantine. Compounds 5, 6 and 7 had δ
values in the range 0.43 to 0.51, which are just slightly higher to that of memantine, 0.39. Based on their δ
values, we can suggest these compounds may have a binding site midway through the pore, maybe a little
deeper but overlapping with that of memantine. This is consistent with binding adjacent to the Q/R/N-site
that determines ion selectivity in ionotropic glutamate receptors.
Cell Toxicity Assay. Compounds 5-7, presenting appreciable NMDAR blocking properties, were selected for
deepening their antioxidant profile in SH-SY5Y human neuroblastoma cells. To this aim, we assessed the
cytotoxicity of compounds 5-7 to define the concentration range to be used in cellular experimental
settings. The antioxidant FA was used for comparison. Cells were exposed to the compounds at
concentrations ranging from 1 to 50 μM for 24 h and cell viability was determined by MTT assay. As shown
in Figure 3, all the compounds were devoid of any toxicity at a concentration up to 20 μM, while only the
shorter derivative 5 retained, like FA, good tolerability up to 50 µM. Lack of toxicity was verified also for
8

ACCEPTED MANUSCRIPT
compound 4, carrying a three methylene spacer, at all the concentrations investigated (data not shown),
confirming that the spacer length significantly influenced compound tolerability in favor of shorter
derivatives.
100
***
75
CTR
50
5
6
7
25
****
FA
****
0
1
5
10
20
50
Conc [microM]
Figure 3. Cellular toxicity of hybrid compounds (5-7) and Ferulic Acid (FA) on human neuroblastoma SH-
SY5Y. Cells were treated with compounds for 24h at different concentrations ranging from 1 to 50μM. Cell
viability was assessed by MTT assay. Data are expressed as percentage of cell viability versus CTR; ***p <
0.001, and ****p < 0.0001 versus CTR; Dunnett’s multiple comparison test, n=3.
Protective Effect toward H2O2-Induced Damage. To determine the antioxidant efficacy of compounds 5-7,
we first studied their ROS scavenging activity when coincubated with 300 μM H2O2, using FA for
comparison. The scavenger effect was evaluated in SH-SY5Y cells by using the fluorescent probe
dichlorodihydrofluorescein diacetate (DCFH-DA) as a marker for quantitative intracellular ROS formation.
The DCFH-fluorescence intensity significantly increased in H2O2-treated cells (black line, Figure 4A) with
respect to untreated cells (dashed grey line, Figure 4A). All compounds, at a concentration of 10 µM, were
able to markedly reduce H2O2-induced intracellular ROS formation, being, however, less effective than FA.
To assess if indirect antioxidant effects could accompany radical scavenger properties, further experiments
were performed pretreating SH-SY5Y cells with compounds 5-7 (10 µM) for 24 h before adding 300 μM
H2O2 (Figure 4B). Again, compounds 5-7 produced a significant reduction in DCHF-fluorescent intensity,
9

ACCEPTED MANUSCRIPT
albeit an inversion in the trend of efficacy was observed. Indeed, with this experimental setting, compound
7 emerged as the most efficacious, reaching FA ability to counteract H2O2-induced ROS formation. Based on
these results, we could speculate that, at least for 7, antioxidant properties might derive from both direct
and indirect effects.
Figure 4. Hybrid compounds reverse ROS formation induced by H2O2. (A) After the loading with DCFH-DA,
SH-SY5Y cells were exposed to 10μM compounds or FA in combination with 300μM of H2O2. The
fluorescence intensity for all compounds tested is significant at any time starting from 60 to 300 min with p
< 0.0001 versus H2O2. At 30 min the significance versus H2O2 is p < 0.01 for compound 7, p < 0.001 for
compound 6 and p < 0.0001 for compound 5 and FA. Dunnett’s multiple comparison test. (B) SH-SY5Y cells
were pre-treated with 10μM of each compound for 24 hours, loaded with DCFH-DA and then exposed to
300μM H2O2. Fluorescence intensity for all compounds tested is significant at any time from 30 to 90 min
with p < 0.0001 versus H2O2. At time 10 min, the fluorescence intensity did not reach statistical significance
for compound 5, whereas for 6 the significance is p < 0.01 and for 7 and FA is p < 0.0001 vs H2O2. At 20 min,
the significance is p < 0.001 for compound 5 and p < 0.0001 for compound 7 and AF vs H2O2. Dunnett’s
multiple comparison test.
Activation of Nrf2 Pathway in SH-SY5Y Cells. The nuclear factor (erythroid-derived 2)-like 2 (Nrf2)
transcriptional pathway is a major player of inducible antioxidant defense [21]. Activation of the Nrf2
pathway, and the subsequent transcription of downstream cytoprotective genes, is triggered by the
disruption of interaction and binding of Nrf2 with the cytosolic Nrf2 repressor Kelch-like ECH-associated
protein 1 (Keap 1) [22]. A variety of electrophiles from synthetic or natural sources is emerging for their
ability to hamper this interaction by targeting key cysteine residues of Keap1, which act as sensors of
oxidative insults [23]. In particular, the electrophilic motif recurring in FA and its derivatives, namely the
α,β-unsaturated carbonyl group, has already been shown to trigger the Nrf2-driven transcriptional process
10

ACCEPTED MANUSCRIPT
in a series of hydroxy-cinnamic derivatives for which trapping Keap1 through covalent adduct formation
was proposed to be the initiating event [24,25]. Thus, we studied compounds 5-7 and FA in SH-SY5Y
neuroblastoma cells to verify whether they may affect the Nrf2 pathway and indirect mechanisms could
therefore contribute to their overall antioxidant profile. To this aim, we first assessed their ability to
modulate the mRNA levels of Nrf2 by real-time PCR, using 10 μM of each compound incubated for 6 h.
Notably, only compound 7 determined a significant increase in Nrf2 mRNA expression (Figure 5, panel A),
while cells treated with FA or compounds 5 and 6 behaved like untreated cells. Coherently with these
results, the same trend was observed when we investigated the ability of compounds to tune the mRNA
levels of heme oxygenase-1 (HO-1), a prototypical Nrf2-target gene related to oxidative stress response.
Indeed, mRNA levels of the inducible cytoprotective gene raised to about 150% of control following
pretreatment with 10 µM 7, while no effect was elicited by FA or compounds 5 and 6 (Figure 5, panel B).
Figure 5. Hybrid compounds modulate Nrf2 and HO-1 mRNA levels. RNA was obtained from cellular
extracts of SH-SY5Y cells treated for 6h with compounds 5-7 and FA at 10 μM and analyzed for Nrf2 (A) and
HO-1 (B) mRNA expression by RT-PCR. GAPDH was used as housekeeping gene. Results are shown as mean
± SEM; *p < 0.05, **p < 0.01 versus CTR; Dunnett’s multiple comparison test, n=3.
Then, we sought to verify whether the increase in HO-1 mRNA expression determined by compound 7
could effectively result in enhanced HO-1 protein levels. To this aim, HO-1 induction was analyzed by
means of Western immunoblotting in the same cell line after treatment for 24 h with 7 at 10 or 20 µM.
Interestingly, compound 7 caused a dose-dependent increase of HO-1 expression, with cells treated with 20
11

ACCEPTED MANUSCRIPT
µM 7 almost doubling HO-1 protein levels of control (Figure 6). These results confirm that compound 7 is a
multimodal antioxidant, which combines radical scavenging properties to the ability of potentiating the
Nrf2/HO-1 axis. Further, the lack of indirect antioxidant efficacy verified for compounds 5, 6 and FA, all
carrying the α,β-unsaturated carbonyl feature, reveal that an electrophilic moiety is not per se sufficient for
activating redox sensor proteins, and shape complementarity may play a pivotal role in this respect.
Particularly, we might speculate that compounds 5-7, varying in the linker length, and FA could differently
orient their cysteine-reactive group toward nucleophilic traps of Keap1 affecting target recognition and,
consequently, a compound’s reactivity and specificity.
24h
250
***
200
150
100
50
0
Figure 6. Effect of compound 7 on HO-1 protein levels. Cellular extracts of SH-SY5Y cells treated for 24 h
with compound 7 at 10 or 20μM were analyzed for HO-1 protein levels by Western Blot. Anti-tubulin was
used as protein loading control. Results are shown as ratio (% of CTR) ± SEM; ***p < 0.001 versus CTR;
Dunnett’s multiple comparison test, n=3.
APP processing in H4-SW cells. In AD, a direct link exists between eNMDAR overactivation and increased
neuronal Aβ production [26]. NMDAR have been proposed to modulate α-secretase activity, shifting APP
metabolism towards a non-amyloidogenic pathway. Memantine has been shown to lower Aβ synthesis in a
number of studies [27,28]. Mechanisms potentially involved in memantine-driven Aβ modulation are not
completely clear, and both NMDAR-mediated and NMDAR-independent mechanisms have been proposed
[29]. In this context, we sought to investigate whether the most promising compounds 5-7 could affect the
APP processing favoring the production of the non-amyloidogenic soluble amyloid precursor protein α
12

ACCEPTED MANUSCRIPT
(sAPPα). Current research suggests that sAPPα plays a role in synaptic growth and plasticity, featuring
neuroprotective and neurotrophic properties [30]. Thus, we studied the effect of the compounds on sAPPα
levels in the human H4 cell line expressing the Swedish mutant form of APP (H4-SW), after 24 h treatment.
Compounds 5-7 were tested at 10 µM concentration, which had no impact on cell viability, as confirmed by
a dose-response curve where memantine, FA and compounds 5-7 had no toxic effect up to 20 µM
concentration (data not shown). Memantine and FA alone were used for comparison. The Western blot
analysis reported in Figure 7 shows that compound 7, but not compounds 5 and 6, significantly increased
sAPPα levels (Figure 7, panel A).
Figure 7. Assessment of the effect of compounds 5-7, MEM and FA on amyloid precursor protein (APP)
proteolytic processing in H4-SW cells. (A) Determination of sAPPα levels. Cells were treated with 10 µM
of each compound and after 24 hours conditioned media were collected and sAPPα content assessed by
Western blotting. The graph shows the densitometric quantification of the Western blotting bands,
normalized to the total protein content of plated cells. *=p<0.05, one-way ANOVA and post-hoc test. (B)
Determination of full-length amyloid precursor protein (flAPP) expression. Cells were treated with 10 µM
of each compound. After 24 hours, H4-SW were lysed and flAPP expression assessed by Western blotting.
The graph shows the densitometric quantification of the Western blotting bands, normalized to α-tubulin
as internal reference.
Notably, compounds 5-7 were not able to affect full-length APP (fl-APP) expression levels, which was
determined in the same cells to evaluate the effect of the compounds on total intracellular APP (Figure 7,
panel B). By enhancing sAPPα levels without affecting the levels of total intracellular APP, compound 7
seems to stimulate APP processing towards the α-secretase (non-amyloidogenic) pathway, which should
result in decreased Aβ production. Noteworthy, at the same concentration, memantine and FA alone
13

ACCEPTED MANUSCRIPT
showed no effect on APP processing, strengthening the value of the design of a hybrid molecule. The lack of
efficacy of memantine, whose potency as NMDAR antagonist is 3-fold higher than that of 7, suggests that
the effect of 7 on APP processing we observed in this experimental setting seems to be not principally
mediated by NMDAR. Interestingly, lengthening of the linker between the pharmacophoric functions up to
six methylenes switched on the efficacy toward both APP processing and Nrf2 activation, pointing to 7 as
the most promising molecule of the series.
Conclusions
NMDAR play a crucial role in the pathophysiology of AD. Excessive activation of NMDAR can compromise
synapse function by triggering neurotoxic events, which involve Aβ peptide and oxidative stress. By
preferentially blocking extrasynaptic rather than synaptic currents, the anti-AD drug memantine limits
neurotoxicity mediated by excessive NMDAR activity while relatively sparing physiological
neurotransmission. This peculiar NMDAR profile prompted us to conjugate memantine with the bioactive
payload FA, aiming to synergistically modulate the critical partnership occurring between oxidative
damage, Aβ burden, and hyperfunctioning NMDAR. For compounds 4-7, which exploit memantine’s
nitrogen for FA connection, chain lengthening positively influenced NMDAR blocking properties. The longer
derivative 7, carrying a hexamethylene spacer between the pharmacophoric functions, presented a
micromolar profile as NMDAR antagonist (IC50=6.9 µM), being only three times less effective than the
parent compound memantine (IC50= 2.3 µM). Further, compound 7 also shares with memantine the binding
site midway through the pore and a voltage-dependent behavior, suggesting that conjugation with FA
produced only a modest perturbation of memantine’s NMDAR binding mode. Compounds with appreciable
NMDAR blocking properties were studied in SH-SY5Y cells to assess their antioxidant properties. All
compounds tested showed notable free radical scavenging effects. Conversely, only 7 was able to
significantly potentiate the expression of Nrf2 and its downstream protective gene HO-1 at the
concentration of 10 µM, therefore emerging as a multimodal antioxidant. Notably, the lack of indirect
14

ACCEPTED MANUSCRIPT
antioxidant efficacy observed for 5 and 6, varying in the linker length, and FA, suggests the importance of
target recognition as a pre-requisite for electrophile reactivity, excluding an indiscriminate effect driven by
the α,β-unsaturated carbonyl group. At the same concentration (10 µM), compound 7, and not shorter
derivatives 5 and 6, significantly enhanced sAPPα levels in H4-SW cells, suggesting that it may stimulate APP
processing in favor of the α-secretase (non-amyloidogenic) pathway and consequently limit Aβ formation.
Thus, the most potent NMDAR antagonist 7 was also able to activate inducible protective pathways which
play a crucial role in contrasting the neurotoxic cascade driven by eNMDAR overactivation. The multimodal
profile of compound 7 was well balanced, in the micromolar-range, and not accompanied by any
cytotoxicity in both SH-SY5Y and H4-SW cells up to the concentration of 20 µM. Based on these findings,
compound 7 emerges as a promising pharmacologic tool for deepening our insight on the significance of
NMDAR-mediated neurotoxic events involving ROS formation and Aβ damage.
Experimental section
Chemistry. Chemical reagents were purchased from Sigma-Aldrich, Fluka and Lancaster (Italy) and used
without further purification. Chromatographic separations were performed on silica gel columns (Kieselgel
40, 0.040-0.063 mm, Merck) by chromatography. Reactions were followed by TLC on Merck (0.25 mm)
glass-packed precoated silica gel plates (60 F254), then visualized with an UV lamp, bromocresol green or
KMnO4. Melting points were measured in glass capillary tubes on a Büchi SMP-20 apparatus and are
uncorrected. Microwave assisted synthesis was performed by using CEM Discover® SP apparatus (2.45 GHz,
maximum power of 300W). NMR spectra were recorded at 400 MHz for 1H and 100 MHz for 13C on Varian
VXR 400 spectrometer. Chemical shifts (δ) are reported in parts per millions (ppm) relative to
tetramethylsilane (TMS), and spin multiplicities are given as s (singlet), br s (broad singlet), d (doublet), t
(triplet), q (quartet), or m (multiplet). Direct infusion ESI-MS mass spectra were recorded on a Waters ZQ
4000 apparatus. Final compounds 1-7 were >95% pure as determined by HPLC analyses. The analyses were
performed under reversed-phase conditions on a Phenomenex Jupiter C18 (150 × 4.6 mm I.D.) column,
15

ACCEPTED MANUSCRIPT
using a binary mixture of 0.1% TFA in H2O/acetonitrile (70/30,v/v for 5; 65/35, v/v for 3, 4, 6, 7; 80/20, v/v
for 1, 2) as the mobile phase, UV detection at λ= 302 nm and a flow rate of 1 mL/min. Analyses were
performed on a liquid chromatograph model PU-2089Plus UV equipped with a 20 μL loop valve (Jasco
Europe, Italy). Compounds were named relying on the naming algorithm developed by CambridgeSoft
Corporation and used in Chem-BioDrawUltra 15.1.
General procedure for the intermediates 11-13. To an ice-cooled solution of ferulic acid (FA, 1 equiv) in dry
DMF (3-4 mL) were added HOBt (1.3 equiv) and EDC (1.3 equiv). The reaction mixture was stirred for 10
min, followed by addition of Et3N (1.3 equiv) and the appropriate mono-protected diamine (8-10) (1 equiv).
Stirring was then continued at room temperature overnight, and the solvent evaporated under vacuum.
The crude was purified by flash chromatography on silica gel using dichloromethane/methanol (9.5:0.5) as
mobile phase.
tert-butyl (E)-(2-(3-(4-hydroxy-3-methoxyphenyl)acrylamido)ethyl)carbamate (11). Synthesized from FA
(400 mg, 2.06 mmol) and 8 [17](330 mg, 2.06 mmol) to afford 11 as waxy solid: 200 mg (30%); 1H NMR (400
MHz, CDCl3) δ 7.50 (d, J = 15.6 Hz, 1H), 6.99 (d, J = 8 Hz, 1H), 6.94 (s, 1H), 6.88 (d, J = 8.4 Hz, 1H), 6.60 (br s,
1H), 6.26 (d, J = 15.6 Hz, 1H), 5.33 (br s, 1H), 3.85 (s, 3H), 3.50-3.46 (m, 2H), 3.33-3.30 (m, 2H), 1.42 (s, 9H).
tert-butyl (E)-(3-(3-(4-hydroxy-3-methoxyphenyl)acrylamido)propyl)carbamate (12). Synthesized from FA
(100 mg, 0.51 mmol) and 9 [16](174 mg, 0.51 mmol) to afford 12 as waxy solid: 100 mg (56%); 1H NMR (400
MHz, CDCl3) δ 7.55 (d, J = 15.2 Hz, 1H), 7.05 (d, J = 8 Hz, 1H), 7.01 (s, 1H), 6.90 (d, J = 8 Hz, 1H), 6.50 (br s,
1H), 6.30 (d, J = 15.2 Hz, 1H), 4.96 (br s, 1H), 3.92 (s, 3H), 3.44-3.40 (m, 2H), 3.23-3.19 (m, 2H), 1.67-1.61 (m,
2H), 1.45 (s, 9H).
tert-butyl (E)-(4-(3-(4-hydroxy-3-methoxyphenyl)acrylamido)butyl)carbamate (13). Synthesized from FA
(400 mg, 2.06 mmol) and 10 [31](188 mg, 2.06 mmol) to afford 13 as waxy green solid: 230 mg (33%); 1H
NMR (400 MHz, CDCl3) δ 7.52 (d, J = 15.2 Hz, 1H), 7.03 (d, J = 8.4 Hz, 1H), 6.98 (s, 1H), 6.88 (d, J = 8 Hz, 1H),
6.25 (d, J = 15.2 Hz, 1H), 6.01 (br s, 1H), 4.64 (br s, 1H), 3.89 (s, 3H), 3.39-3.38 (m, 2H), 3.17-3.13 (m, 2H),
1.60-1.55 (m, 4H), 1.43 (s, 9H).
16

ACCEPTED MANUSCRIPT
General procedure for the intermediates 14-16. To an ice-cooled solution of the appropriate Boc-protected
intermediate (11-13, 1 equiv) in CH2Cl2 (2-3 mL) was added HCl 4 M in dioxane (2-3 mL) and the reaction
mixture was stirred at 0 °C for 90 min. The solvent was evaporated and the crude purified by flash
chromatography on silica gel using dichloromethane/methanol/aqueous ammonia 33% (8:2:0.2) affording
desired intermediates as free bases.
(E)-N-(2-aminoethyl)-3-(4-hydroxy-3-methoxyphenyl)acrylamide (14). Synthesized from 11 (200 mg, 0.60
mmol) to afford 14 as pale yellow solid: 120 mg (86%); 1H NMR (400 MHz, DMSO-d6) δ 8.34 (t, J = 5.4 Hz,
1H), 7.72 (d, J = 15.6 Hz, 1H), 7.53 (s, 1H), 7.41-7.38 (m, 1H), 7.20 (d, J = 7.6 Hz, 1H), 6.86 (d, J = 15.6 Hz, 1H),
4.21 (s, 3H), 3.59-3.55 (m, 2H), 3.04-3.00 (m, 2H), 2.92 (br s, 2H).
(E)-N-(3-aminopropyl)-3-(4-hydroxy-3-methoxyphenyl)acrylamide (15). Synthesized from 12 (100 mg, 0.30
mmol) to afford 15 as pale green solid: 71 mg (99%); 1H NMR (400 MHz, CD3OD) δ 7.40 (d, J = 15.6 Hz, 1H),
7.10 (s, 1H), 7.00-6.98 (m, 1H), 6.77 (d, J = 8.4 Hz, 1H), 6.48 (d, J = 15.6 Hz, 1H), 3.84 (s, 3H), 3.38 (t, J = 6.4
Hz, 2H), 2.96 (t, J = 7.2 Hz, 2H), 1.95-1.88 (m, 2H).
(E)-N-(4-aminobutyl)-3-(4-hydroxy-3-methoxyphenyl)acrylamide (16). Synthesized from 13 (230 mg, 0.63
mmol) to afford 16 as pale green solid: 160 mg (96%); 1H NMR (400 MHz, CD3OD) δ 7.39 (d, J = 15.6 Hz, 1H),
7.05 (s, 1H), 6.96 (d, J = 7.6 Hz, 1H), 6.73 (d, J = 8.4 Hz, 1H), 6.40 (d, J = 15.6 Hz, 1H), 3.82 (s, 3H), 3.31-3.23
(m, 2H), 2.80-2.78 (m, 2H), 1.59-1.58 (m, 4H).
General procedure for compounds 1-3. To an ice-cooled solution of the hydrochloride salt 17 [18](1 equiv)
in dry DMF (3 mL) were added HOBt (1.3 equiv) and EDC (1.3 equiv) under N2 atmosphere. The reaction
mixture was stirred for 10 min, followed by addition of the appropriate intermediates (14-16) (2 equiv).
Stirring was continued at room temperature for 36-48 h, and then the solvent evaporated under vacuum.
The crude was purified by column chromatography on silica gel using dichloromethane/methanol/aqueous
ammonia 33% (8.5:1.5:0.15) as mobile phase.
(1r,3s,5R,7S)-3-amino-N-(2-((E)-3-(4-hydroxy-3-methoxyphenyl)acrylamido)ethyl)-5,7-
dimethyladamantane-1-carboxamide (1). Synthesized from 17 (64 mg, 0.24 mmol) and 14 (115 mg, 0.48
17

ACCEPTED MANUSCRIPT
mmol) to afford 1 as green solid: 80 mg (74%); mp 122-124 °C; 1H NMR (400 MHz, DMSO-d6) δ 7.99 (br s,
1H), 7.46 (br s, 1H), 7.32 (d, J = 15.6 Hz, 1H), 7.11 (s, 1H), 6.99-6.97 (m, 1H), 6.80 (d, J = 8.4 Hz, 1H), 6.40 (d, J
= 15.6 Hz, 1H), 3.80 (s, 3H), 3.20-3.17 (m, 2H), 3.15-3.12 (m, 2H), 1.40 (s, 2H), 1.29-1.23 (m, 4H), 1.16-1.12
(m, 4H), 1.01 (s, 2H), 0.82 (s, 6H). 13C NMR (100 MHz, DMSO-d6) δ 176.77, 166.17, 148.84, 148.27, 139.54,
126.70, 121.97, 119.25, 116.10, 111.23, 55.96, 51.30, 49.90, 49.79, 49.03, 46.18, 44.55, 44.10, 38.90, 32.95,
30.19. MS [ESI+] m/z 442 [M+1]+.
(1r,3s,5R,7S)-3-amino-N-(3-((E)-3-(4-hydroxy-3-methoxyphenyl)acrylamido)propyl)-5,7-
dimethyladamantane-1-carboxamide (2). Synthesized from 17 (37 mg, 0.14 mmol) and 15 (71 mg, 0.28
mmol) to afford 2 as green solid: 31 mg (52%); mp 118-119 °C; 1H NMR (400 MHz, CD3OD) δ 7.43 (d, J = 16
Hz, 1H), 7.11 (s, 1H), 7.03-7.00 (m, 1H), 6.78 (d, J = 8 Hz, 1H), 6.41 (d, J = 16 Hz, 1H), 3.87 (s, 3H), 3.29 (t, J =
8 Hz, 2H), 3.22 (t, J = 8 Hz, 2H), 1.73-1.66 (m, 2H), 1.63 (s, 2H), 1.44 (s, 4H), 1.32-1.26 (m, 4H), 1.14 (s, 2H),
0.92 (s, 6H). 13C NMR (100 MHz, CD3OD) δ 177.79, 168.00, 148.87, 148.00, 140.82, 126.55, 121.91, 117.08,
115.16, 110.09, 54.95, 50.48, 48.96, 48.41, 48.28, 44.09, 43.78, 43.43, 36.40, 36.37, 32.47, 28.98, 28.60. MS
[ESI+] m/z 455 [M+1]+.
(1r,3s,5R,7S)-3-amino-N-(4-((E)-3-(4-hydroxy-3-methoxyphenyl)acrylamido)butyl)-5,7-
dimethyladamantane-1-carboxamide (3). Synthesized from 17 (89 mg, 0.30 mmol) and 16 (160 mg, 0.61
mmol) to afford 3 as green solid: 68 mg (48%); mp 116-117 °C; 1H NMR (400 MHz, CD3OD) δ 7.40 (d, J = 15.6
Hz, 1H), 7.05 (s, 1H), 6.98-6.95 (m, 1H), 6.73 (d, J = 8 Hz, 1H), 6.37 (d, J = 15.6 Hz, 1H), 3.82 (s, 3H), 3.27-3.24
(m, 2H), 3.18-3.15 (m, 2H), 1.54-1.49 (m, 6H), 1.38-1.36 (m, 4H), 1.24 (s, 4H), 1.07 (s, 2H), 0.86 (s, 6H). 13C
NMR (100 MHz, CD3OD) δ 177.94, 167.83, 149.83, 148.29, 140.76, 126.04, 122.02, 116.86, 115.46, 110.03,
54.92, 49.76, 49.12, 49.09, 48.47, 44.22, 44.09, 43.91, 38.77, 38.63, 32.46, 28.76, 26.53, 26.44. MS [ESI+]
m/z 470 [M+1]+.
General procedure for the intermediates 22-25. A mixture of memantine hydrochloride (MEM, 1 equiv),
K2CO3 (2 equiv), KI (1 equiv) and the appropriate intermediate (18-21, 1 equiv) in dry DMF (2-5 ml) was
placed in a microwave (140°C, 250 Psi, 100 W) and left stirring for 1 h. The solvent was removed under
18

ACCEPTED MANUSCRIPT
reduced pressure and the crude purified by chromatography on silica gel using
dichloromethane/methanol/aqueous ammonia 33% (9:1:0.2) as mobile phase.
tert-butyl (3-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)propyl)carbamate (22). Synthesized from
18 [14](0.4 g, 1.2 mmol) to afford 22 as a pale oil: 0.24 g (59%); 1H NMR (400 MHz, CDCl3) δ 5.34 (br s, 1H),
3.16-3.14 (m, 2H), 2.65 (t, J = 6.6 Hz, 2H), 2.09-2.07 (m, 1H), 1.67-1.64 (m, 2H), 1.49 (s, 2H), 1.38 (s, 9H),
1.32-1.23 (m, 8H), 1.06-1.05 (m, 2H), 0.78 (s, 6H).
tert-butyl (4-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)butyl)carbamate (23). Synthesized from
19 [32](0.4 g, 1.5 mmol) to afford 23 as a pale oil: 0.35 g (87%); 1H NMR (400 MHz, CDCl3) δ 5.32 (br s, 1H),
3.00-2.99 (m, 2H), 2.63 (t, J = 7.2 Hz, 2H), 2.06-2.04 (m, 1H), 1.59-1.55 (m, 4H), 1.48-1.43 (m, 2H), 1.37-1.28
(m, 13H), 1.24-1.13 (m, 4H), 1.02-1.01 (m, 2H), 0.73 (s, 6H).
tert-butyl (5-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)pentyl)carbamate (24). Synthesized from
20 [33](0.4 g, 1 mmol) to afford 24 as a pale oil: 0.22 g (55%); 1H NMR (400 MHz, CDCl3) δ 5.31 (br s, 1H),
3.11-3.09 (m, 2H), 2.59 (t, J = 7.4 Hz, 2H), 2.15-2.13 (m, 1H), 1.54-1.42 (m, 15H), 1.40-1.25 (m, 10H), 1.16-
1.11 (m, 2H), 0.85 (s, 6H).
tert-butyl (6-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)hexyl)carbamate (25). Synthesized from
21 [34](0.6 g, 1.6 mmol) to afford 25 as a pale oil: 0.4 g (79%); 1H NMR (400 MHz, CDCl3) δ 4.61 (br s, 1H),
2.99-2.97 (m, 2H), 2.48 (t, J = 7.4 Hz, 2H), 2.03-2.02 (m, 1H), 1.41-1.32 (m, 15H), 1.24-1.15 (m, 12H), 1.01-
1.00 (m, 2H), 0.73 (s, 6H).
General procedure for the intermediates 26-29. To an ice-cooled appropriate Boc-protected intermediate
(22-25, 1 equiv) was added HCl 4 M in dioxane (4 mL) and the reaction mixture was stirred at 0 °C for 15-20
min under N2 atmosphere. The solvent was evaporated and the crude purified by flash chromatography on
silica gel using dichloromethane/methanol/aqueous ammonia 33% (8:2:0.4) affording desired
intermediates as free bases.
N1-((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)propane-1,3-diamine (26). Synthesized from 22 (0.24 g, 0.7
mmol) to afford 26 as a pale oil: 0.09 g (54%); 1H NMR (400 MHz, DMSO-d6) δ 3.94 (br s, 3H), 2.74-2.73 (m,
19

ACCEPTED MANUSCRIPT
2H) 2.59 (t, J = 6.8 Hz, 2H), 2.06-2.04 (m, 1H), 1.57- 1.50 (m, 2H), 1.42-1.41 (m, 2H), 1.26-1.18 (m, 8H), 1.09-
1.00 (m, 2H), 0.78 (s, 6H).
N1-((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)butane-1,4-diamine (27). Synthesized from 23 (0.35 g, 1
mmol) to afford 27 as a pale oil: 0.18 g (72%); 1H NMR (400 MHz, CDCl3) δ 2.66-2.62 (m, 5H), 2.53 (t, J = 6.8
Hz, 2H), 2.06-2.03 (m, 1H), 1.48-1.40 (m, 6H), 1.26-1.17 (m, 8H), 1.05-0.98 (m, 2H), 0.74 (s, 6H).
N1-((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)pentane-1,5-diamine (28). Synthesized from 24 (0.22 g, 0.6
mmol) to afford 28 as a pale oil: 0.11 g (70%); 1H NMR (400 MHz, CDCl3) δ 2.63 (t, J = 7.1 Hz, 2H), 2.52 (t, J =
7.1 Hz, 2H), 2.07-2.06 (m, 1H), 1.59 (br s, 3H), 1.42-1.38 (m, 6H), 1.31-1.17 (m, 10H), 1.05-1.04 (m, 2H), 0.78
(s, 6H).
N1-((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)hexane-1,6-diamine (29). Synthesized from 25 (0.4 g, 1.06
mmol) to afford 29 as a pale oil: 0.16 g (54%); 1H NMR (400 MHz, CDCl3) δ 2.63 (t, J = 7.2 Hz, 2H), 2.52 (t, J =
7.2 Hz, 2H), 2.38 (br s, 3H), 2.08-2.07 (m, 1H), 1.44-1.38 (m, 6H), 1.30-1.19 (m, 12H), 1.06-1.05 (m, 2H), 0.78
(s, 6H).
General procedure for compounds 4-7. To an ice-cooled solution of FA (1 equiv) in dry DMF (3-4 mL) was
added HOBt (1.3 equiv) and EDC (1.3 equiv) under N2 atmosphere. The reaction mixture was stirred for 10
min at 0°C, followed by addition of Et3N (1.3 equiv) and the appropriate amine (26-29) (1 equiv). Stirring
was then continued at room temperature overnight, then the solvent was evaporated under reduced
pressure and the crude purified by chromatography on silica gel.
(E)-N-(3-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)propyl)-3-(4-hydroxy-3-
methoxyphenyl)acrylamide (4). Synthesized from 26 (90 mg, 0.38 mmol), purified by chromatography on
silica gel using petroleum ether/dichloromethane/methanol/aqueous ammonia 33% (2:6.5:1.5:0.16) as
mobile phase to afford 4 as a yellow solid: 61.5 mg (39%); mp 213-215 °C; 1H NMR (400 MHz, CDCl3) δ 7.66
(br s, 1H), 7.49 (d, J = 15.6 Hz, 1H), 7.09 (s, 1H), 7.08-7.04 (m, 1H), 6.88 (d, J = 8 Hz, 1H), 6.64 (d, J = 15.6 Hz,
1H), 3.90 (s, 3H), 3.57-3.52 (m, 2H), 2.94-2.91 (m, 2H), 2.24 (m, 3H), 1.89-1.88 (m, 2H), 1.72-1.63 (m, 4H),
1.43-1.32 (m, 4H), 1.20 (s, 2H), 0.85 (s, 6H). 13C NMR (100 MHz, CDCl3) δ 167.60, 147.53, 147.02, 140.33,
20

ACCEPTED MANUSCRIPT
127.35, 122.15, 119.07, 114.87, 110.05, 59.04, 56.03, 49.88, 44.16, 41.92, 39.60, 38.59, 37.00, 32.57, 29.67,
26.77, 24.06. MS [ESI+] m/z 413 [M+1]+.
(E)-N-(4-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)butyl)-3-(4-hydroxy-3-
methoxyphenyl)acrylamide (5). Synthesized from 27 (72.5 mg, 0.28 mmol), purified by chromatography on
silica gel using petroleum ether/dichloromethane/methanol/aqueous ammonia 33% (2:6.5:1.5:0.07) as
mobile phase to afford 5 as a yellow solid: 39.6 mg (32%); mp 174-175 °C; 1H NMR (400 MHz, CDCl3) δ 7.83
(br s, 1H), 7.41 (d, J = 15.6 Hz, 1H), 7.04 (s, 1H), 6.95 (d, J = 8.2 Hz, 1H), 6.86 (d, J = 8.2 Hz, 1H), 6.70 (d, J =
15.6 Hz, 1H), 3.81 (s, 3H), 3.33-3.32 (m, 2H), 2.82 (t, J = 7.4 Hz, 2H), 2.17 (s, 1H), 1.89-1.88 (m, 4H), 1.71-
1.62 (m, 4H), 1.37-1.23 (m, 6H), 1.14 (s, 2H), 0.84 (s, 6H). 13C NMR (100 MHz, CDCl3) δ 167.60, 147.53,
147.02, 140.33, 127.35, 122.15, 119.07, 114.87, 110.05, 59.04, 56.03, 49.88, 44.16, 41.92, 39.60, 38.59,
37.00, 32.57, 29.67, 26.77, 24.06. MS [ESI+] m/z 427 [M+1]+.
(E)-N-(5-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)pentyl)-3-(4-hydroxy-3-
methoxyphenyl)acrylamide (6). Synthesized from 28 (54 mg, 0.2 mmol), purified by chromatography on
silica gel using petroleum ether/dichloromethane/methanol/aqueous ammonia 33% (2:6.5:1.5:0.09) as
mobile phase to afford 6 as a yellow solid: 27.7 mg (31%); mp 200-202 °C; 1H NMR (400 MHz, CDCl3) δ 7.51
(d, J = 15.6 Hz, 1H), 7.03 (d, J = 7.6 Hz, 1H), 6.96 (s, 1H), 6.86 (d, J = 7.6 Hz, 1H), 6.31 (d, J = 15.6 Hz, 1H), 5.82
(br s, 1H), 3.86 (s, 3H), 3.37-3.33 (m, 2H), 2.61 (t, J = 7.2 Hz, 2H), 2.13-2.11 (m, 1H), 1.58-1.53 (m, 4H), 1.36-
1.24 (m, 12H), 1.10-1.09 (m, 2H), 0.82 (s, 6H). 13C NMR (100 MHz, CDCl3) δ 166.21, 147.43, 146.81, 140.72,
127.33, 121.93, 118.40, 114.84, 109.72, 55.86, 50.83, 48.43, 42.89, 40.83, 40.37, 39.52, 32.37, 30.24, 30.19,
29.45, 24.78. MS [ESI+] m/z 441 [M+1]+.
(E)-N-(6-(((1r,3R,5S,7r)-3,5-dimethyladamantan-1-yl)amino)hexyl)-3-(4-hydroxy-3-
methoxyphenyl)acrylamide (7). Synthesized from 29 (0.1 g, 0.36 mmol), purified by chromatography on
silica gel using petroleum ether/dichloromethane/methanol/aqueous ammonia 33% (2:6.5:1.5:0.1) as
mobile phase to afford 7 as a yellow solid: 95.9 mg (59%); mp 203-204°C; 1H NMR (400 MHz, CDCl3) δ 7.51
(d, J = 15.2 Hz, 1H), 7.03 (d, J = 8.4 Hz, 1H), 6.97 (s, 1H), 6.86 (d, J = 8.4 Hz, 1H), 6.25 (d, J = 15.2 Hz, 1H), 5.73
21

ACCEPTED MANUSCRIPT
(br s, 1H), 3.88 (s, 3H), 3.36-3.31 (m, 2H), 2.57 (t, J = 7.4 Hz, 2H), 2.12-2.09 (m, 1H), 1.54-1.45 (m, 6H), 1.33-
1.23 (12H), 1.13-1.05 (m, 2H), 0.82 (s, 6H). 13C NMR (100 MHz, CDCl3) δ 166.20, 147.57, 146.90, 140.74,
127.26, 121.96, 118.35, 114.90, 109.75, 55.85, 50.85, 48.42, 42.90, 40.79, 40.45, 39.54, 32.37, 30.24, 30.20,
29.50, 27.03, 26.73. MS [ESI+] m/z 455 [M+1]+.
Electrophysiological assays. Inhibition of NMDARs by compounds 1-7 and memantine was assessed by the
expression of GluN1-1a and GluN2A subunits in Xenopus oocytes followed by voltage clamp recording.
Oocytes were obtained from the European Xenopus Resource Centre (University of Portsmouth, UK)
directly following their removal from mature female Xenopus laevis according to UK Home Office
guidelines. Sections of the ovary were cut into smaller pieces and treated with 1 mg/mL collagenase type
1A (Sigma-Aldrich) in Ca2+-free modified Barth’s media containing 96 mM NaCl, 2 mM KCl, 1 mM MgCl2, 5
mM HEPES, 2.5 mM pyruvic acid, 0.5 mM theophylline, 0.05 mg/mL gentamicin, pH 7.5, with shaking at 18
°C for 40-60 minutes, in order to separate them into individual defolliculated oocytes. The oocytes were
then rinsed in Ca2+-free modified Barth’s media multiple times until the solution was clear and stored in
modified Barth’s media (as per Ca2+-free but including 1.8 mM CaCl2). Oocytes were injected with 50 nL of
cRNA encoding both the GluN1-1a and GluN2A subunits (1:1 by weight ratio; total 250 ng/μL). The cRNA
was synthesized from linearized plasmid DNA (pRK7) containing the GluN-encoding genes using an
mMessage mMachine kit (Invitrogen). Following injection, oocytes were kept in modified Barth’s media at
18 °C for 3-4 days before electrophysiological testing. Voltage-clamp recording was conducted using an
Axoclamp-2A voltage-clamp amplifier (Axon Instruments, USA). Microelectrodes were pulled from
borosilicate glass capillaries (TW150F-4, World Precision Instruments) using a Sutter P-97 programmable
micropipette puller to have a resistance of 0.5-2 MΩ when filled with 3 M KCl. Oocytes were placed in a
perfusion chamber and constantly perfused ( 5 mL/min) with solution containing 96 mM NaCl, 2 mM KCl,
1.8 mM CaCl2, 10 mM HEPES, pH 7.5, and voltage-clamped at holding potentials (Vh) between −40 and −100
mV. NMDAR currents were initiated by application of 100 μM NMDA + 10 μM glycine. Once the current had
reached a steady state ( 30 s) the test compounds were introduced at concentrations ranging from 0.01 to
100 μM until a new plateau (inhibited current) was achieved. All agonists and test compounds were applied
22

ACCEPTED MANUSCRIPT
using an Automate Valvelink 8 perfusion system. Analogue output from the amplifier was digitized by a CED
1401 plus A/D converter at 100 Hz and recorded on a windows PC using WinEDR software (Dr John
Dempster, University of Strathclyde, UK). NMDA/glycine-evoked current in the presence of test compound
was normalized to that in its absence just before test compound addition (% control response) and plotted
against concentration. Concentration-inhibition data were fit by:
ꢉꢊꢊ
% ꢀꢁꢂꢃꢄꢁꢅ ꢄꢆꢇꢈꢁꢂꢇꢆ = ꢉꢋ ꢉꢊ
Equation 1
((ꢌꢍꢎ ꢑꢒ ꢔ ꢕ) × ꢖꢗꢘꢘꢙꢘꢚꢛꢜ)
ꢏꢐ ꢓꢐ
to obtain IC50 values, where X = Log10[compound]; using Graphpad Prism 7. All points were means of at
least 5 separate oocytes. For voltage dependence studies the test compounds were applied at a single
( IC50) concentration but at four Vhs in the range -40 to -100 mV. Data were normalized as above and fit by:
ꢉꢊꢊ
% ꢀꢁꢂꢃꢄꢁꢅ ꢄꢆꢇꢈꢁꢂꢇꢆ =
Equation 2
ꢨꢩ
ꢦꢧ(
ꢞ ꢠ
ꢟ
ꢣ
)
ꢪꢫ
ꢝꢉꢋ
ꢣ
ꢤꢥ
ꢢ
ꢡ
to obtain δ values (fraction of the membrane electric field crossed by the blocking compound), where [B] is
the concentration of the blocker, KD is the dissociation constant at 0 mV, z is the charge valence of the
blocker, F is Faraday’s constant, E is the membrane potential, R is the universal gas constant and T is
absolute temperature; using Graphpad Prism 7. All points were means of at least 5 separate oocytes.
Reagents for cellular experiments (SH-SY5Y cells). All hybrid compounds were solubilized in DMSO (at
stock concentrations) and frozen (−20°C) in aliquots that were diluted immediately prior to use. For each
experimental setting, one stock aliquot was thawed out and diluted to minimize compound damage due to
repeated freeze and thaw cycles. The final concentration of DMSO in culture medium was <0.1%. Ferulic
Acid was purchased from Sigma Aldrich (Merck KGaA, Darmstadt, Germany). Rabbit polyclonal anti-human
HO-1 (NBP1-31341) antibody was purchased from Novus (Biotechne, Minneapolis USA). Mouse monoclonal
anti-β-tubulin (T0198) was purchased from by Sigma Aldrich (Merck KGaA, Darmstadt, Germany).
Cell cultures. All culture media, supplements and Foetal Bovine Serum (FBS) were purchased from Sigma
Aldrich (Merck KGaA, Darmstadt, Germany). Human neuroblastoma SH-SY5Y cells from the European
Collection of Cell Cultures (ECACC No. 94030304) were cultured in a medium with equal amounts of Eagle’s
23

ACCEPTED MANUSCRIPT
minimum essential medium and Nutrient Mixture Ham’s F-12, supplemented with 10% heat-inactivated
FBS, 2 mM glutamine, 0.1 mg/mL streptomycin, 100 IU·mL penicillin and non-essential aminoacids at 37 °C
in 5% CO2 and 95% air atmosphere. H4-SW cells were cultured in D-MEM medium (Invitrogen, Carlsbad, CA)
supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin and 2mM
L-glutamine. Hygromycin B and Blasticidin S were used as selection antibiotics for SW mutation
maintenance.
Cell Viability. The mitochondrial dehydrogenase activity that reduces 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyl-tetrazolium bromide (MTT, Sigma Aldrich, Merck KGaA, Darmstadt, Germany) was used to
determine cellular viability, in a quantitative colorimetric assay. At day 0, SH-SY5Y cells were plated at a
density of 2.5x104 viable cells per well in 96-well plates. After treatment, according to the experimental
setting, cells were exposed to an MTT solution in complete medium (1 mg/mL). Following 4 h incubation
with MTT and treatment with sodium dodecyl sulfate (SDS) for 24 h, cell viability reduction was quantified
by using a Synergy HT multidetection microplate reader (Bio-Tek).
Measurement of Intracellular ROS. DCFH-DA (Sigma Aldrich, Merck KGaA, Darmstadt, Germany) was used
to estimate intracellular ROS following two different experimental setting described in each figure legend.
In each setting, cells were loaded with 25 μM DCFH-DA for 45 min. After centrifugation DCFH-DA was
removed, and the results were visualized using Synergy HT multidetection microplate reader (BioTek) with
excitation and emission wavelengths of 485 and 530 nm, respectively.
Real-time PCR (RT-PCR). For RNA extraction, 2x106 cells were used. Total RNA was extracted using a Direct-
zolTM RNA MiniPrep (Zymo Research, Irvine, USA) following the manufacturer’s instructions. QuantiTect
reversion transcription kit and QuantiTect Sybr Green PCR kit (Qiagen, Valencia, CA, USA) were used for
cDNA synthesis and gene expression analysis, following the manufacturer’s specifications. Nrf2, HO-1, and
GAPDH primers were provided by Qiagen. GAPDH was used as an endogenous reference.
Immunodetection of HO-1, flAPP and sAPPα. The expression of HO-1 in whole cell lysates was assessed by
Western blot analysis. After treatment, cell monolayers were washed twice with ice-cold PBS, lysed on the
24

ACCEPTED MANUSCRIPT
culture dish by the addition of ice-cold homogenization buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM
EDTA, 0.5% Triton X-100 and protease inhibitor mix) and an aliquot was used for protein quantification,
whereas the remainder was prepared for western blot by mixing the cell lysate with 2X sample buffer (125
mM Tris-HCl pH 6.8, 4% SDS, 20% glycerol, 6% β-mercaptoethanol, 0.1% bromophenol blue) and then
denaturing at 95°C for 5 min. Equivalent amounts of extracted proteins were loaded into a SDS-PAGE gel,
electrophoresed under reducing conditions, transferred to a PVDF membrane (Sigma Aldrich, Merck KGaA,
Darmstadt, Germany) and then blocked for 1 h with 5% w/v BSA in Tris-buffered saline containing 0.1%
Tween 20 (TBS-T). The proteins were visualized using primary antibodies for HO-1, full length (fl) APP or
soluble APP alpha (sAPPα) and α- or β-tubulin (1:1000) followed by secondary horseradish peroxidase
conjugated antibody (1:5000) diluted in 5% w/v BSA in TBS-T. Tubulins were performed as a control for gel
loading. Signal development was carried out using an enhanced chemiluminescent method (Pierce,
Rockford, IL, USA).
Densitometry and statistics. All experiments, unless specified, were performed at least three times with
representative results being shown. Data are expressed as mean ± SEM. The relative densities of the
acquired images of Western blotting bands were analyzed with ImageJ software. Statistical analyses were
performed using Prism software version 7.0 (GraphPad Software, La Jolla, CA, USA). Statistical differences
were determined by analysis of variance (ANOVA) followed, when significant, by an appropriate post hoc
test as indicated in figure legends. A p value of <0.05 was considered statistically significant.
Acknowledgments: This work was supported by the University of Bologna (grants from the RFO).
25

ACCEPTED MANUSCRIPT
References
1.
2.
3.
Niu H, Álvarez-Álvarez I, Guillén-Grima F, Aguinaga-Ontoso I. Prevalence and incidence of
Alzheimer's disease in Europe: A meta-analysis. Neurologia, 32(8), 523-532 (2017).
Traynelis SF, Wollmuth LP, McBain CJ et al. Glutamate receptor ion channels: structure, regulation,
and function. Pharmacol Rev, 62(3), 405-496 (2010).
Müller MK, Jacobi E, Sakimura K, Malinow R, von Engelhardt J. NMDA receptors mediate synaptic
depression, but not spine loss in the dentate gyrus of adult amyloid Beta (Aβ) overexpressing mice.
Acta Neuropathol Commun, 6(1), 110 (2018).
4.
Hardingham GE, Fukunaga Y, Bading H. Extrasynaptic NMDARs oppose synaptic NMDARs by
triggering CREB shut-off and cell death pathways. Nat Neurosci, 5(5), 405-414 (2002).
Lipton SA. NMDA receptor activity regulates transcription of antioxidant pathways. Nat Neurosci,
11(4), 381-382 (2008).
5.
6.
Lanni C, Fagiani F, Racchi M et al. Beta-amyloid short- and long-term synaptic entanglement.
Pharmacol Res, 139, 243-260 (2019).
7.
Tu S, Okamoto S, Lipton SA, Xu H. Oligomeric Aβ-induced synaptic dysfunction in Alzheimer's
disease. Mol Neurodegener, 9, 48 (2014).
8.
Xia P, Chen HS, Zhang D, Lipton SA. Memantine preferentially blocks extrasynaptic over synaptic
NMDA receptor currents in hippocampal autapses. J Neurosci, 30(33), 11246-11250 (2010).
Song X, Jensen M, Jogini V et al. Mechanism of NMDA receptor channel block by MK-801 and
memantine. Nature, 556(7702), 515-519 (2018).
9.
10.
11.
Takahashi H, Xia P, Cui J et al. Pharmacologically targeted NMDA receptor antagonism by
NitroMemantine for cerebrovascular disease. Sci Rep, 5, 14781 (2015).
Figueiredo CP, Clarke JR, Ledo JH et al. Memantine rescues transient cognitive impairment caused
by high-molecular-weight aβ oligomers but not the persistent impairment induced by low-
molecular-weight oligomers. J Neurosci, 33(23), 9626-9634 (2013).
12.
Benchekroun M, Romero A, Egea J et al. The Antioxidant Additive Approach for Alzheimer's Disease
Therapy: New Ferulic (Lipoic) Acid Plus Melatonin Modified Tacrines as Cholinesterases Inhibitors,
Direct Antioxidants, and Nuclear Factor (Erythroid-Derived 2)-Like 2 Activators. J Med Chem, 59(21),
9967-9973 (2016).
13.
14.
Lipton SA. Paradigm shift in neuroprotection by NMDA receptor blockade: memantine and beyond.
Nat Rev Drug Discov, 5(2), 160-170 (2006).
Simoni E, Daniele S, Bottegoni G et al. Combining galantamine and memantine in multitargeted,
new chemical entities potentially useful in Alzheimer's disease. J Med Chem, 55(22), 9708-9721
(2012).
15.
Reggiani AM, Simoni E, Caporaso R et al. In Vivo Characterization of ARN14140, a
Memantine/Galantamine-Based Multi-Target Compound for Alzheimer's Disease. Sci Rep, 6, 33172
(2016).
16.
17.
18.
19.
Minniti E, Byl JAW, Riccardi L et al. Novel xanthone-polyamine conjugates as catalytic inhibitors of
human topoisomerase IIα. Bioorg Med Chem Lett, 27(20), 4687-4693 (2017).
Riva E, Comi D, Borrelli S et al. Synthesis and biological evaluation of new camptothecin derivatives
obtained by modification of position 20. Bioorg Med Chem, 18(24), 8660-8668 (2010).
Wanka L, Cabrele C, Vanejews M, Schreiner P. gamma-aminoadamantanecarboxylic acids through
direct C-H bond amidations. European Journal of Organic Chemistry, (9), 1474-1490 (2007).
Sobolevsky A, Koshelev S, Khodorov B. Probing of NMDA channels with fast blockers. Journal of
Neuroscience, 19(24), 10611-10626 (1999).
20.
21.
Woodhull AM. Ionic blockage of sodium channels in nerve. J Gen Physiol, 61(6), 687-708 (1973).
Cuadrado A, Rojo AI, Wells G et al. Therapeutic targeting of the NRF2 and KEAP1 partnership in
chronic diseases. Nat Rev Drug Discov, 18(4), 295-317 (2019).
22.
Dinkova-Kostova AT, Kostov RV, Canning P. Keap1, the cysteine-based mammalian intracellular
sensor for electrophiles and oxidants. Arch Biochem Biophys, 617, 84-93 (2017).
26

ACCEPTED MANUSCRIPT
23.
24.
25.
26.
Basagni F, Lanni C, Minarini A, Rosini M. Lights and shadows of electrophile signaling: focus on the
Nrf2-Keap1 pathway. Future Med Chem, 11(7), 707-721 (2019).
Simoni E, Serafini MM, Caporaso R et al. Targeting the Nrf2/Amyloid-Beta Liaison in Alzheimer's
Disease: A Rational Approach. ACS Chem Neurosci, 8(7), 1618-1627 (2017).
Simoni E, Bergamini C, Fato R et al. Polyamine conjugation of curcumin analogues toward the
discovery of mitochondria-directed neuroprotective agents. J Med Chem, 53(19), 7264-7268 (2010).
Bordji K, Becerril-Ortega J, Nicole O, Buisson A. Activation of extrasynaptic, but not synaptic, NMDA
receptors modifies amyloid precursor protein expression pattern and increases amyloid-ß
production. J Neurosci, 30(47), 15927-15942 (2010).
27.
28.
29.
30.
31.
Folch J, Busquets O, Ettcheto M et al. Memantine for the Treatment of Dementia: A Review on its
Current and Future Applications. J Alzheimers Dis, 62(3), 1223-1240 (2018).
Alley GM, Bailey JA, Chen D et al. Memantine lowers amyloid-beta peptide levels in neuronal
cultures and in APP/PS1 transgenic mice. J Neurosci Res, 88(1), 143-154 (2010).
Bordji K, Becerril-Ortega J, Buisson A. Synapses, NMDA receptor activity and neuronal Aβ
production in Alzheimer's disease. Rev Neurosci, 22(3), 285-294 (2011).
Habib A, Sawmiller D, Tan J. Restoring Soluble Amyloid Precursor Protein α Functions as a Potential
Treatment for Alzheimer's Disease. J Neurosci Res, 95(4), 973-991 (2017).
Palermo G, Minniti E, Greco M et al. An optimized polyamine moiety boosts the potency of human
type II topoisomerase poisons as quantified by comparative analysis centered on the clinical
candidate F14512. Chemical Communications, 51(76), 14310-14313 (2015).
Nguyen C, Ruda GF, Schipani A et al. Acyclic nucleoside analogues as inhibitors of Plasmodium
falciparum dUTPase. J Med Chem, 49(14), 4183-4195 (2006).
Lamanna G, Smulski C, Chekkat N et al. Multimerization of an Apoptogenic TRAIL-Mimicking
Peptide by Using Adamantane-Based Dendrons. Chemistry-a European Journal, 19(5), 1762-1768
(2013).
32.
33.
34.
Zanichelli V, Bazzoni M, Arduini A et al. Redox-Switchable Calix[6]arene-Based Isomeric Rotaxanes.
Chemistry, 24(47), 12370-12382 (2018).
27

ACCEPTED MANUSCRIPT
Highlights
•
•
•
•
•
Memantine targets NMDAR overactivation which leads to oxidative stress and Aβ damage
We conjugated memantine with the antioxidant ferulic acid, affording hybrids 1-7
Compound 7 is, like memantine, a micromolar open channel blocker of NMDAR
It also exerts direct and indirect antioxidant effects, and increases sAPPα levels
7 is a valuable tool to study NMDAR-mediated neurotoxic events involving ROS and Aβ