Identification of a neuroprotective and selective butyrylcholinesterase inhibitor derived from the natural alkaloid evodiamine

Article abstract of DOI:10.1016/j.ejmech.2014.05.002

Two sets of carbamates based on the natural alkaloid evodiamine were desined, synthesized and evaluated as potential butyrylcholinesterase inhibitors. Althouh a set of carbamates of 3-hydroxyevodiamine (10a-f) is inactive both at AChE and BChE, carbamates of 5-deoxo-3-hydroxyevodiamine (11a-f) exhibit much better potency with selectivity toward BChE. The heptyl carbamate of 5-deoxo-3-hydroxyevodiamine (11c) shows the best potency with an IC₅₀ value of 77 nM and very ood selectivity over AChE. ORAC and cell-based assays indicate 11c owns pronounced antioxidant properties with 1.75 Trolox equivalents and stron neuroprotection even from 1 μM onwards. These combined activities miht enable compound 11c to be a potential candidate for treatment of Alzheimer's disease.

Full text of DOI:10.1016/j.ejmech.2014.05.002

European Journal of Medicinal Chemistry 81 (2014) 15e21
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Short communication
Identification of a neuroprotective and selective butyrylcholinesterase
inhibitor derived from the natural alkaloid evodiamine
,
Guozheng Huang a, Beata Kling b, Fouad H. Darras b, Jörg Heilmann b, Michael Decker a
*
a Pharmaceutical and Medicinal Chemistry, Institute of Pharmacy and Food Chemistry, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany
b Pharmaceutical Biology, Institute of Pharmacy, University of Regensburg, Universitätsstraße 31, D-93053 Regensburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Two sets of carbamates based on the natural alkaloid evodiamine were designed, synthesized and
Received 22 December 2013
Received in revised form
7 April 2014
Accepted 1 May 2014
Available online 4 May 2014
evaluated as potential butyrylcholinesterase inhibitors. Although
a set of carbamates of 3-
hydroxyevodiamine (10aef) is inactive both at AChE and BChE, carbamates of 5-deoxo-3-
hydroxyevodiamine (11aef) exhibit much better potency with selectivity toward BChE. The heptyl
carbamate of 5-deoxo-3-hydroxyevodiamine (11c) shows the best potency with an IC50 value of 77 nM
and very good selectivity over AChE. ORAC and cell-based assays indicate 11c owns pronounced anti-
oxidant properties with 1.75 Trolox equivalents and strong neuroprotection even from 1 mM onwards.
These combined activities might enable compound 11c to be a potential candidate for treatment of
Alzheimer’s disease.
Keywords:
Alzheimer’s disease
Butyrylcholinesterase inhibitor
Neuroprotection
Evodiamine
Ó 2014 Elsevier Masson SAS. All rights reserved.
Carbamates
1. Introduction
understood, it may be also involved in the normal nervous system
function and might participate in pathological processes in the
Alzheimer’s disease (AD), the most prominent form of dementia,
is clinically defined as a progressive and irreversible loss of cogni-
tive functions combined with neuropsychiatric disturbance.
Although numerous efforts have been dedicated to the investiga-
tion of its etiologies, no effective method is yet available to prevent
or reverse AD and its pathogenesis still remains unclear [1]. Several
hypotheses have been advocated to constitute the key pathological
etiology of AD as elucidated in several studies [6]. During later stage
of AD, the activity of AChE decreases to reach 10e15% of normal
values in certain brain regions, while the amount of BChE remains
the same or even increases up to 2-fold [7]. Hence, selective AChE
inhibitors might therefore encounter clinical ineffectiveness in later
stages of AD. Actually, a balance between AChE and BChE inhibition
or selectivity toward BChE might be more desirable. Furthermore,
in vivo studies proved selective BChE inhibitors elevate brain
hallmarks of AD, including b-amyloid deposits, s-protein aggrega-
tion, oxidative stress, and cholinergic system dysfunction [2]. Ac-
cording to the cholinergic hypothesis, drastic deficit of the
neurotransmitter acetylcholine is the ultimate reason for failure in
neurotransmission and subsequent aggravating cognitive and
memory symptoms. Therefore, acetylcholinesterase (AChE) in-
hibitors are utilized clinically for the treatment of mild AD, such as
tacrine, donepezil, rivastigmine and galantamine [3]. Current me-
dicinal chemistry research focuses on novel AChE inhibitors, often
with multifunctional and multi-target activities [4,5].
acetylcholine, lower b-amyloid peptide, and ameliorate cognitive
dysfunction [8,9]. In summary, selective BChE inhibitors could be of
importance for treatment of AD and exploration of the pathogen-
esis of AD.
Up to now, only a few selective BChE inhibitors (with selectivity
index, SI > 100) were developed, such as heterobivalent tacrine
derivatives [10], benzofurans [11] and isosorbide-based compounds
[12] (the most potent one is isosorbide-2-benzyl carbamate-5-
salicylate, 1), cymserine analogs [13,14] (N1-phenylethyl-norcyser-
ine, 2), and diarylimidazole compounds [15] (3). Our group used
quinazolinimines as lead structures to develop a series of N-
bridgeheaded tri- and tetracylic compounds with improved selec-
tivity toward BChE [16e18]. Further exploration on the related
structures identified a carbamate-based tetracyclic lead structure
(4) [19]. Performing SARs on a library of tri- and tetracyclic struc-
tures with and without an additional carbamate moiety revealed
The second type of cholinesterase in the human body, butyr-
ylcholinesterase (BChE), is also known as plasma cholinesterase.
Although its portfolio of functions is still not completely
* Corresponding author.
E-mail address: Michael.Decker@uni-wuerzburg.de (M. Decker).
http://dx.doi.org/10.1016/j.ejmech.2014.05.002
0223-5234/Ó 2014 Elsevier Masson SAS. All rights reserved.

16
G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
that a tetracyclic structure holds better potency and selectivity than
the tricyclic structure. This result intrigued us to explore the pos-
sibility of better interaction between the aromatic ring and the
binding region of BChE.
It is well established that oxidative stress, the formation of
reactive oxygen species (ROS), and subsequent neurotoxicity are
key processes in the pathophysiology of AD [20]. We showed pre-
viously that compound 4 and its analogs possess very significant
antioxidant properties (determined both with regard to their
physico-chemical radical scavenging properties as well as with
regard to neuron protection against oxidative stress) [19].
Evodiamine is the major quinazoline alkaloid isolated from
Evodia rutaecarpa used in Traditional Chinese Medicine for treat-
ment of a variety of ailments including headache, abdominal pain,
postpartum hemorrhage, dysentery and amenorrhea [21,22].
Several research efforts indicate that evodiamine and its chemical
derivatives are potent anti-cancer drug candidates [23,24]. The
physiological effects of evodiamine on aged animals implicate a
potential for the treatment of erectile dysfunction [25]. An in vivo
study proved that it also can improve cognitive abilities in a
transgenic mouse model of AD [26]. Recently, Lee et al. described a
specific inhibitory activity on BChE with an IC50 value of 1.7
(corresponding to 5.6 M) although this value is far lower than for
the compounds shown in Chart 1 [27].
mg/mL
m
In this study, we wanted to extend and continue our earlier
studies seeking highly selective and potent BChE inhibitors with
pronounced neuroprotective effects by combining the potent nat-
ural compound evodiamine with a privileged carbamate scaffold
and applying the knowhow we had gained from previous tetracy-
clic scaffolds.
2. Results and discussion
2.1. Chemistry
Fig. 2. Evaluation of neurotoxicity (A) and neuroprotection (B) of compounds 14, 9, 10c
and 11c, respectively, against glutamate induced oxidative stress on HT-22 cells. Re-
sults are presented as means ꢀ SD and refer to untreated control cells which were set
as 100% values. Experiments were carried out with 4 parallels and repeated inde-
pendently at least three times. Data were subjected to one-way ANOVA followed by
Dunnet’s multiple comparison post-test using GraphPad Prism 5 Software. Levels of
significance are: p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***).
Condensation of 4,9-dihydro-3H-pyrido[3,4-b]indole (6) and 6-
(benzyloxy)-1-methyl-1H-benzo[d] [1,3]oxazine-2,4-dione (5)
in refluxing dichloromethane afforded the desired 3-
benzyloxyevodiamine (7) in excellent yield (97%) (Scheme 1).
Intermediate 6 could be obtained from indole via a six-step linear
synthesis. The isatoic anhydride derivative 5, which represents an
activated form of anthranilic acid with a benzylated phenolic OH,
was synthesized in five steps with almost quantitative overall yield
starting from 5-hydroxyanthranilic acid (cf. Supplementary Mate-
rial for the syntheses of the two intermediates). Benzylation of the
phenolic group in compound 5 is necessary to avoid methylation of
the phenolic group and to improve solubility in dichloromethane
for subsequent fusion reaction. 3-Benzyloxyevodiamine (7) was
Fig. 1. (A) Time-dependent pattern of BChE inhibition by compound 11c (0.25e1 mM)
and (B) Stability constants of the inhibitoreBChE complex (KC) and rate constants of
carbamoyl-ChE formation (k3) of 11c and physostigmine, respectively.
Chart 1. Structures of selective BChE inhibitors.

G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
17
Scheme 1. Synthetic protocol for target carbamates 10aef and 11aef. Reagents and conditions: (i) CH2Cl2, reflux, 20 h; (ii) 1 atm H2, THF, r.t., 20 h; (iii) LiAlH4, THF, r.t., 8 h; (iv)
isocyanate, Hünig’s base, THF, 2e4 days; (v) isocyanate, triethylamine, CH2Cl2, 2e4 h.
debenzylated by hydrogenation to convert it into 3-
hydroxyevodiamine (8). Carbamoylation of 3-hydroxyevodiamine
with appropriate isocyanates in the presence of Hünig’s base
generated the target carbamates 10aef [19]. 3-Hydoxyevodiamine
was reduced by lithium aluminum hydride to give 5-deoxo-3-
hydroxyevodiamine (9), which could also be carbamoylated using
isocyanates in the presence of triethylamine to afford the second
type of carbamates (11aef) (Scheme 1). Evodiamine (13) and 5-
deoxoevodiamine (14) were synthesized according to a recent
literature method [23] (Scheme 2). The compounds’ structures and
purities were fully supported by analytical and spectral data (1H
NMR, 13C NMR, HPLC, and HRMS).
respective benzo analog we previously described [19]. Neverthe-
less, reduction of 8 to 5-deoxo-3-hydroxyevodiamine 9 recovered
potency, with moderate selectivity toward BChE.
Surprisingly, carbamates of 3-hydroxyevodiamine (10aef) are
not inhibiting either AChE or BChE at 10 mM, while after carba-
moylation, the set of 5-deoxo-3-hydroxyevodiamine derivatives
(11aef) are exhibiting better potency than 5-deoxo-3-
hydroxyevodiamine 9. Aromatic carbamates (11def) are less
active with IC50 values higher than 10
concentrations could not be determined because of insolubility at
>10 M). Aliphatic carbamates (11aec) are more potent though,
mM (inhibition at higher
m
showing IC50 values in the micromolar and submicromolar range.
The by far most potent and most selective BChE inhibitor is the
heptyl carbamate of 5-deoxo-3-hydroxyevodiamine (11c), which
represents a 2-digit nanomolar inhibitor with no significant activity
2.2. Pharmacology
on AChE even at 10 mM. Its potency toward BChE was as good as the
2.2.1. Inhibition assay of AChE and BChE
positive control physostigmine, but exhibiting pronounced BChE
selectivity.
Synthesized carbamates and synthetic intermediates, together
with evodiamine and 5-deoxoevodiamine, were investigated for
inhibition of the hydrolase activity of AChE and BChE using Ellman’s
reaction [19,28]. The results are summarized in Table 1.
Our lead structure evodiamine (13) is moderately potent toward
2.2.2. Kinetic study of BChE inhibition
Rivastigmine (ExelonÒ) and the natural product physostigmine
represent pseudo-irreversible inhibitors transferring the carba-
mate moiety to the serine unit in the active center of AChE and
BChE (pseudo-irreversible, since the carbamate subsequently
slowly hydrolyzes off the enzyme). Based on the structural simi-
larity of 3-hydroxyevodiamine 11c and physostigmine, we assumed
the 11c would own similar pseudo-irreversible inhibitory behavior.
In order to obtain more detailed information about the time course
BChE with an IC50 of 4.62
mM and shows moderate selectivity,
which is consistent with literature [27]. Reduction of evodiamine to
14 resulted in higher potency and better selectivity which indicates
removal of 5-position’s amide oxygen is well tolerated for main-
taining potency. In contrast to our expectation, introduction of 3-
hydroxy group into evodiamine rendered compound 8 to lose ac-
tivity on both AChE and BChE. This is in strong contrast to the
Scheme 2. Synthetic protocol for evodiamine (13) and 5-deoxoevodiamine (14) synthesis. Reagents and conditions: (i) CH2Cl2, reflux, 20 h; (ii) LiAlH4, THF, r.t., 8 h.

18
G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
Table 1
AChE and BChE inhibition, selectivity index and antioxidant capacities expressed as Trolox equivalents (concentration range 1e5 mM).
Test compound
IC50
,
m
M (pIC50 ꢀ SEM or inhibition in % at 10
m
M)
SIb
Trolox equ.
AChEa
BChEa
Physostigmine
0.0077 (8.11 ꢀ 0.08)
3.01 (5.52 ꢀ 0.06)
>100 (29%)
0.060 (7.22 ꢀ 0.06)
0.880 (6.07 ꢀ 0.05)
0.13
3.42
3-Benzyloxyevodiamine (7)
3-Hydroxyevodiamine (8)
5-Deoxo-3-hydroxyevodiamine (9)
R ¼ ethyl (10a)
>100 (47% at 100
19.3 (4.71 ꢀ 0.05)
>10 (21%)
mM)
>100 (43% at 100
>10 (15%)
m
M)
>5.17
3.55 ꢀ 0.03
0.32 ꢀ 0.03
R ¼ butyl (10b)
>10 (10%)
>10 (13%)
>10 (12%)
>10 (1%)
R ¼ heptyl (10c)
>10 (10%)
>10 (14%)
>10 (15%)
>10 (7%)
(10d)
(10e)
>10 (15%)
>10 (3%)
(10f)
R ¼ ethyl (11a)
R ¼ butyl (11b)
R ¼ heptyl (11c)
>10 (26%)
>10 (47%)
>10 (17%)
1.03 (5.99 ꢀ 0.06)
3.97 (5.40 ꢀ 0.12)
0.077 (7.11 ꢀ 0.06
>9.71
>2.52
>>130
1.75 ꢀ 0.14
>10 (34%)
>10 (13%)
2.15 (5.67 ꢀ 0.10)
>10 (41%)
>4.65
(11d)
(11e)
>10 (19%)
>10 (30%)
(11f)
Evodiamine (13)
5-Deoxoevodiamine (14)
>10 (37%)
>10 (24%)
4.62 (5.34 ꢀ 0.10)
2.68 (5.57 ꢀ 0.09)
>2.16
>3.73
2.32 ꢀ 0.24
a
AChE from electric eel and BChE from equine serum; data are the means of at least 3 independent determinations.
Selectivity index [IC50(AChE)/IC50(BChE)].
b
of inhibition and therefore more precise data about the equilibrium
carbamate transfer to the enzyme; therefore this potent antioxi-
dant is the neuroprotectant to act in vivo. Surprisingly, 5-
deoxoevodiamine 14, without a free hydroxy group is also potent
with 2.32 Trolox equivalents. Finally, carbamate 10c is almost
inactive in the ORAC assay with only 0.32 Trolox equivalents
(Table 1). All the indolo compounds exhibit significantly higher
Trolox values compared to the previously described benzo com-
pounds, this is especially true for the unsubstituted compound 5-
deoxoevodiamine 14, which shows the importance of the indole
ring with regard to antioxidant properties [19].
constant of inhibition, a kinetic study of compound 11c (and
physostigmine as positive control) was performed on BChE (Fig. 1)
[19]. Compound 11c shows a time-dependent pattern of inhibition
characterized by an increase until a steady state after 60 min. The
equilibrium constant of the inhibitoreBChE complex (KC) and the
carbamoylation rate constant (k3) of 11c were determined (Fig. 1).
The heptyl carbamate 11c represents a highly potent pseudo-
irreversible BChE inhibitor with KC ¼ 492 nM. The rate constant
of 0.243 minꢁ1 shows rapid carbamoylation.
2.2.3. Antioxidant capacities (ORAC assay)
2.2.4. Neuroprotectivity
For evaluation of antioxidant potencies, in a first assay the target
compounds’ radical scavenging capacities were determined by
means of their ability to reduce the amount of peroxyl radicals
(oxygen radical absorbance capacity assay, ORAC assay). The most
potent compound 11c and its free hydroxy group containing
starting material 9 were chosen. Also 5-deoxoevodiamine (14) was
tested for comparison to screen the influence of the free hydroxy
group. Compound 10c was tested to evaluate the influence of the 5-
position amide’s oxygen. As expected, 11c is potent with antioxi-
dant capacity of 1.75 Trolox equivalents, meaning that it is more
potent than the positive control Trolox, a water-soluble vitamin E
derivative. 5-Deoxo-3-hydroxyevodiamine 9 with a free hydroxy
group is much more potent with 3.55 Trolox equivalents, which
indicates a free hydroxy group is favorable for antioxidant potency.
The hydroxyl compound 9 is formed from compound 11c after
To obtain physiologically relevant data of antioxidant com-
pounds with regard to neuroprotection, the neuronal cell line HT-
22 was chosen as a model system, which is a glutamate sensitive
cell line and derived from murine hippocampal tissue [29]. As these
cells lack ionotropic glutamate receptors, addition of extracellular
glutamate in high concentrations inhibits cystine (as oxidized form
of cysteine) transport via the cystine/glutamate antiporter. This
neuronal oxidative stress induced toxicity leads to intracellular
cysteine and therefore glutathione depletion which induces intra-
cellular ROS accumulation and cell injuries [30e32].
The same compounds selected for the ORAC assay (9, 10c, 11c,
and 14) were subsequently evaluated at different concentrations
ranging from 1 to 25 mM for their neuroprotective potential toward
neuronal HT-22 cells exposed to glutamate and subsequent intra-
cellular ROS generation and oxidative stress. Furthermore, the

G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
19
compounds were tested for putative neurotoxic effects toward the
HT-22 neurons to elucidate possible self-toxic effects toward the
cells. Concerning potential neurotoxicity, neither compound 9 with
its free hydroxy group nor the corresponding carbamate 11c
significantly reduce the cells’ viability at concentrations 1, 5 or
High-resolution ESI mass (HRMS) spectral data were acquired on an
Agilent 1100 series LC/MSD ion trap mass spectrometer and a Agilent
6520B UPLC-TOF-MS instrument (Agilent, Santa Clara, U.S.A.),
respectively. Analytic HPLC was performed on a system from Shi-
madzu Products equipped with a DGU-20A3R controller, LC20AB
liquid chromatograph, and a SPD-20A UV/Vis detector. Stationary
phase was a Synergi 4U fusion-RP (150 ꢂ 4.6 mm) column. As mobile
phase, a gradient MeOH-TFA (0.01%)/water-TFA (0.01%) (phase A/
phase B) were used. Gradient mode: 0e8 min (30e90% phase A), 8e
13 min (90% phase A), 15e18 min (90e30% phase A).
Intermediates 5, 6, 12, evodiamine (13), and 5-deoxoevodiamine
(14), respectively, were synthesized according to literature pro-
cedures describing individual steps, but not in the required
sequence. Therefore, sequence and synthetic details can be found in
the Supplementary Material.
10
25
m
m
M but significantly reduce the viability of HT-22 neurons at
M each (Fig. 2A). The same observation can be made for 5-
deoxoevodiamine 14. In contrast to these compounds, the carba-
mate 10c reveals significant neurotoxic effects by reducing the cells’
viability at all tested concentrations from 1
compound was inactive in the ORAC assay.
mM onward e this
As expected, compound 9 with the free hydroxy group that is
also formed after BChE carbamoylation from compound 11c shows
prominent neuroprotection toward the glutamate challenged HT-
22 cells nearly to the same extent as the very potent positive
control quercetin from 5
is slightly reduced with increasing concentrations, probably due to
the neurotoxic effect which was measurable from 25 M onwards
(Fig. 2A). Again, also 5-deoxoevodiamine 14 showed neuro-
protective effects at 10 and 25 M. Therefore, the physico-chemical
mM onwards (Fig. 2B). The protective effect
4.1. Synthesis of 3-(benzyloxy)-14-methyl-7,8,13b,14-
tetrahydroindolo[20,30:3,4] pyrido[2,1-b]quinazolin-5(13H)-one (3-
benzyloxyevodiamine, 7)
m
m
radical scavenging properties determined in the ORAC assay are
reflected in the neuronal cell-based assay.
It has been shown before that not only compounds with the free
hydroxy group but also the corresponding carbamates can act as
neuroprotectants in their own regards [19]. And indeed evaluation of
the corresponding heptyl carbamate (11c) of compound 9 revealed
A solution of 6-(benzyloxy)-1-methyl-1H-benzo[d][1,3] oxa-
zine-2,4-dione (5, 1.416 g, 5 mmol) and 4,9-dihydro-3H-pyrido[3,4-
b]indole (6, 0.87 g, 5.11 mmol, 1.02 eq.) in CH2Cl2 (30 mL) was
stirred under reflux for 20 h. Then it was concentrated, the residue
was rinsed in diethyl ether to give the product as a yellow solid
(1.93 g, 97%). 1H NMR (400 MHz, CDCl3):
d 8.27 (s, 1H, NH), 7.71 (d,
strong neuroprotective properties even from 1
m
M onwards to the
J ¼ 2.7 Hz, 1H), 7.63e7.57 (m, 1H), 7.49e7.37 (m, 5H), 7.36e7.31 (m,
same extent as compound 9 (Fig. 2B). Additionally, the heptyl
carbamate 10c, which varies only in one oxygen atom from com-
pound 11c, still owns slightly measurable neuroprotective proper-
ties, which might even have been more pronounced without the
compounds self-toxicity toward the cells that was detected and is
shown in Fig. 2A.
In summary, with regard to biological activities, both ORAC and
hippocampal neuronal cell-based assays indicate that the most
potent BChE inhibitor 11c also acts as a potent neuroprotectant. It is
worth to note that although the free hydroxy group is favorable for
neuroprotectivity, an indole ring system containing molecules
without free hydroxy group also remains active in its own regard.
1H), 7.29e7.24 (m, 1H), 7.21e7.07 (m, 3H), 5.88 (t, J ¼ 1.4 Hz, 1H,
C13b-H), 5.16e5.06 (m, 2H, PhCH2O), 4.91e4.84 (m, 1H, C7eH),
3.34e3.26 (m, 1H, C7eH), 3.03e2.88 (m, 2H, C8eH), 2.38 (s, 3H,
NCH3). 13C NMR (101 MHz, CDCl3):
d
164.60 (C5), 155.96, 144.72,
136.89, 136.87, 128.76 (2 ꢂ C), 128.40, 128.21, 127.76 (2 ꢂ C), 126.41,
125.18, 124.78, 123.19, 122.09, 120.17, 119.06, 113.68, 112.22, 111.48,
70.62 (PhCH2O), 69.22 (C13b), 39.55 (C7), 37.52 (NCH3), 20.31 (C8).
HPLC purity ¼ 98.65%, tR ¼ 9.58 min. HRMS (ESI): calcd for [M þ H]þ
(C26H24N3O2) requires m/z 410.1863, found 410.1867.
4.2. Synthesis of 3-hydroxy-14-methyl-7,8,13b,14-tetrahydroindolo
[20,30:3,4] pyrido[2,1-b]quinazolin-5(13H)-one (3-
hydroxyevodiamine, 8)
3. Conclusion
3-Benzyloxyevodiamine (7, 1.85 g, 4.518 mmol) in dried THF
(100 mL) was hydrogenated under 1 atm of hydrogen catalyzed by
500 mg palladium on carbon (10%) at r.t. for 20 h. The catalyst was
filtered off, the filtrate was concentrated to afford the 3-
hydroxyevodiamine as a yellow solid (1.39 g, 97%). 1H NMR
By derivation of the biologically active natural alkaloid evodi-
amine two sets of carbamates were identified with BChE inhibitory
properties. The heptyl carbamate of 5-deoxo-3-hydroxyevodiamine
(11c) showed the best potency with an IC50 value of 77 nM and
excellent selectivity over BChE, representing one of the few com-
pounds with such an activity and selectivity profile. Carbamate 11c
and hydroxy compound 9 formed after carbamate transfer are very
potent neuroprotectants.
(400 MHz, DMSO-d6): d 11.24 (s, 1H, NH), 9.46 (s, 1H, 3-OH), 7.51 (d,
J ¼ 7.8 Hz,1H), 7.37 (d, J ¼ 8.1 Hz,1H), 7.29 (d, J ¼ 2.9 Hz,1H), 7.15e7.09
(m, 1H), 7.07 (d, J ¼ 8.6 Hz, 1H), 7.05e7.00 (m, 1H), 6.96 (dd, J ¼ 8.6,
2.9 Hz, 1H), 5.95 (s, 1H, C13b-H), 4.70e4.56 (m, 1H, C7eH), 3.14e3.18
(m, 1H, C7eH), 2.93e2.86 (m, 1H), 2.85e2.75 (m, 1H, C8eH), 2.37 (s,
4. Experimental
3H, N14eCH3). 13C NMR (101 MHz, DMSO-d6):
d 163.66 (C5), 153.38,
142.58, 136.83, 129.36, 125.73, 123.75, 123.50, 121.87, 120.85, 118.81,
118.36, 113.08, 111.61, 111.57, 68.95 (C13b), 39.38 (C7), 36.95 (NCH3),
19.80 (C8). HPLC purity ¼ 99.71%, tR ¼ 7.73 min. HRMS (ESI): calcd for
[M þ H]þ (C19H18N3O2) requires m/z 320.1394, found 320.1396.
Common reagents and solvents were obtained from commercial
suppliers and used without further purification. Tetrahydrofuran
(THF) was distilled from sodium/benzophenone under argon atmo-
sphere. Reactions were conducted using dried flasks under an at-
mosphere of nitrogen. Reaction progress was monitored using
analytical thin layer chromatography (TLC) on precoated silica gel
GF254 plates (MachereyeNagel GmbH & Co. KG, Düren, Germany)
and spots were detected under UV light (254 nm). Nuclear magnetic
resonance spectra were recorded with a Bruker AV-400 NMR in-
strument (Bruker, Karlsruhe, Germany) in DMSO-d6 or CDCl3.
Chemical shifts are expressed in ppm relative to CDCl3 or DMSO-d6
(7.26/2.50 and 77.16/39.52 ppm for 1H and 13C NMR, respectively).
4.3. Synthesis of 14-methyl-5,7,8,13,13b,14-hexahydroindolo
[20,30:3,4]pyrido[2,1-b] quinazolin-3-ol(3-hydroxy-5-
deoxoevodiamine, 9)
To the cooled suspension of lithium aluminum hydride (0.61 g,
16 mmol, 4 eq.) in dried THF (40 mL) was added a solution of 3-
hydroxyevodiamine (8, 1.29 g, 4 mmol) in dried THF (40 mL)
dropwise at 0 ꢃC. After addition, the mixture was stirred at r.t. for

20
G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
8 h, and then cooled on an ice-bath again. To the cooled mixture
was added NaSO4$10H2O in small portions until no gas was liber-
ated any more. Then it was filtered through a pad of Celite, the cake
washed with CH2Cl2/MeOH (5/1, 5 ꢂ 40 mL). The filtrate was
combined and concentrated in vacuo to give a yellow oil which was
purified by column chromatography (CH2Cl2/MeOH ¼ 10/1) to
afford the product as a yellow solid (1.05 g, 86%). 1H NMR (400 MHz,
(dd, J ¼ 8.7, 2.5 Hz,1H), 6.83 (d, J ¼ 2.5 Hz,1H), 4.99 (t, J ¼ 5.7 Hz,1H,
heptyl-NH), 4.75 (s, 1H, C13b-H), 4.03 (d, J ¼ 15.2 Hz, 1H, C5eH), 3.82
0
(d, J ¼ 15.2 Hz, 1H, C5eH), 3.37e3.21 (m, 3H, C7eH & 2 ꢂ C1 H2),
3.10e2.97 (m, 1H, C7eH), 2.85e2.71 (m, 2H, C8eH), 2.60 (s, 3H,
0
N
14eCH3), 1.62e1.52 (m, 2H, C2 H2), 1.41e1.27 (m, 8H, (CH2)4), 0.89
(t, J ¼ 6.4 Hz, 3H, CH3). 13C NMR (101 MHz, CDCl3)
d 155.06 (CO),
146.07, 145.92, 136.57, 130.86, 127.94, 126.98, 123.56, 122.27, 120.51,
DMSO-d6):
d
11.06 (s, 1H, N13eH), 8.98 (s, 1H, 3-OH), 7.44 (d,
119.87, 119.64, 118.68, 112.08, 111.25, 74.27 (C13b), 56.23 (C5), 50.71
0 0 0 0
(C7), 41.45 (C1 ), 39.78 (NeCH3), 31.88 (C2 ), 29.99 (C3 ), 29.08 (C4 ),
J ¼ 7.8 Hz, 1H), 7.33 (d, J ¼ 8.0 Hz, 1H), 7.10e7.02 (m, 1H), 7.01e6.94
(m, 1H), 6.88 (d, J ¼ 8.7 Hz, 1H), 6.60 (dd, J ¼ 8.6, 2.8 Hz, 1H), 6.47 (d,
J ¼ 2.7 Hz, 1H), 4.64 (s, 1H, C13b-H), 3.93 (d, J ¼ 15.3 Hz, 1H, C5eH),
3.69 (d, J ¼ 15.2 Hz, 1H, C5eH), 3.28e3.21 (m, 1H, C7eH), 2.91e2.80
0
0
26.86 (C5 ), 22.73 (C6 ), 21.21 (C8), 14.21 (CH3). HPLC purity ¼ 99.19%,
þ
tR ¼ 7.88 min. HRMS (ESI): calcd for [M þ H] (C27H35N4O2) re-
quires m/z 447.2755, found 447.2766.
(m, 1H, C7eH), 2.74e2.58 (m, 2H, C8eH), 2.44 (s, 3H, N14eCH3).13
C
NMR (101 MHz, DMSO-d6):
d
152.18, 140.54, 136.68, 131.40, 127.99,
4.6. Inhibition assay of AChE and BChE
126.19, 123.73, 121.04, 118.37, 117.93, 114.29, 112.45, 111.38, 110.10,
74.15 (C13b), 55.87 (C5), 50.03 (C7), 39.56 (N14eCH3), 20.87 (C8).
HPLC purity ¼ 97.94%, tR ¼ 4.53 min. HRMS (ESI): calcd for [M þ H]þ
(C19H20N3O) requires m/z 306.1601, found 306.1605.
AChE (E.C.3.1.1.7, Type VI-S, from Electric Eel) and BChE
(E.C.3.1.1.8, from equine serum) were purchased from Sigmae
Aldrich (Steinheim, Germany). DTNB (Ellman’s reagent), ATC and
BTC iodides were obtained from Fluka (Buchs, Switzerland). The
assay was performed as described in the following procedure [18,
19, and 28]. Stock solutions of the test compounds were prepared in
4.4. General procedure for preparation of carbamates of 3-
hydroxyevodiamine (10aef)
ethanol, 100 m
L of which gave a final concentration of 10ꢁ3 M when
To the solution of 3-hydroxyevodiamine (8, 64 mg, 0.2 mmol) in
dried THF (8 mL) were added diisopropylethylamine (52 mg, 68 mL,
diluted to the final volume of 1.66 mL. The highest concentration of
the test compounds applied in the assay was 10ꢁ4 M (3% EtOH in
the stock solution did not influence enzyme activity). In order to
obtain an inhibition curve, at least five different concentrations
(normally 10ꢁ4e10ꢁ9 M) of the test compound were measured at
25 ꢃC and 412 nm, each concentration in triplicate. For buffer
preparation, 3.12 g of sodium dihydrogen phosphate (20 mmol)
were dissolved in 500 mL of water and adjusted with NaOH to
0.4 mmol, 2.0 eq.) and the appropriate isocyanate (0.4 mmol, 2.0
eq.). The reaction mixture was stirred at r.t for 2e4 days; during
stirring a solid precipitated. Then the mixture was concentrated in
vacuo to remove about 3 mL of THF, the resulting white solid was
collected by filtration and rinsed with diethyl ether to afford the
target compound.
pH
¼
8.0
ꢀ
0.1. Enzyme solutions were prepared to give
4.4.1. 14-Methyl-5-oxo-5,7,8,13,13b,14-hexahydroindolo[20,30:3,4]
2.5 units mLꢁ1 in 1.4 mL aliquots. Furthermore, 0.01 M DTNB so-
lution, 0.075 M ATC and BTC solutions, respectively, were used. A
pyrido[2,1-b]quinazolin-3-yl heptyl carbamate (10c)
White solid (61 mg, 66%). 1H NMR (400 MHz, DMSO-d6)
d
11.14
cuvette containing 1.5 mL of phosphate buffer, 50
respective enzyme, 10 L of DTNB and 50 L of the test compound
solution was allowed to stand for 30 min, and the reaction was
started by addition of 10 L of the substrate solution (ATC/BTC). The
mL of the
(s, 1H, N13eH), 7.72 (t, J ¼ 5.7 Hz, 1H, NHCOO), 7.49 (d, J ¼ 7.8 Hz,
m
m
1H), 7.46 (d, J ¼ 2.8 Hz, 1H), 7.37 (d, J ¼ 8.1 Hz, 1H), 7.24 (dd, J ¼ 8.7,
2.8 Hz, 1H), 7.16e7.09 (m, 2H), 7.05e6.97 (m, 1H), 6.10 (s, 1H, C13b
-
m
H), 4.68e4.59 (m, 1H, C7eH), 3.24e3.16 (m, 1H, C7eH), 3.04 (dd,
solution was mixed immediately, and exactly 4.5 min after sub-
strate addition the absorption was measured. For the reference
0
J ¼ 13.0, 6.7 Hz, 2H, C1 H2), 2.94e2.78 (m, 2H, C8eH), 2.74 (s, 3H,
0
N
14eCH3), 1.53e1.40 (m, 2H, C2 H2), 1.27 (s, 8H, (CH2)4), 0.87 (t,
value, 100
mL of water replaced the test compound solution. For
J ¼ 6.8 Hz, 3H, CH3). 13C NMR (101 MHz, DMSO-d6):
d
163.48 (C5),
determining the blank value, additionally 100 mL of water replaced
154.43 (NHCOO), 146.30, 145.21, 136.59, 129.89, 127.09, 125.84,
the enzyme solution. The inhibition curve was obtained by plotting
the percentage enzyme activity (100% for the reference) versus
logarithm of test compound concentration.
121.89, 120.86, 120.36, 119.94, 118.87, 118.27, 111.63, 111.56, 69.38
0
0
(C13b), 40.44 (2 ꢂ C, C7 and C1 ), 36.57 (N14eCH3), 31.19 (C2 ), 29.15
0
0
0
0
(C3 ), 28.34 (C4 ), 26.16 (C5 ), 22.01 (C6 ),19.54 (C8), 13.92 (CH3). HPLC
purity ¼ 100%, tR ¼ 9.62 min. HRMS (ESI): calcd for [M þ H]þ
(C27H33N4O3) requires m/z 461.2547, found 461.2552.
4.7. ORAC assay
The antioxidant activity was determined by the oxygen radical
absorbance capacityefluorescein (ORAC-FL) assay [33], modified
by Dávalos et al. [34]. The ORAC assay measures antioxidant
scavenging activity against peroxyl radicals, their formation
induced by 2,20-azobis(2-amidinopropane) dihydrochloride
(AAPH) at 37 ꢃC.
4.5. General procedure for preparation of carbamates of 5-deoxo-3-
hydroxyevodiamine (11a-f)
To the solution of 14-methyl-5,7,8,13,13b,14-hexahydroindolo
[20,30:3,4]pyrido[2,1-b]quinazolin-3-ol (61 mg, 0.2 mmol) in
CH2Cl2 (15 mL) were added triethylamine (22 mg, 30.6
mL,
The reaction was carried out in 75 mM phosphate buffer (pH
0.22 mmol, 1.1 eq.) and appropriate isocyanate (0.22 mmol, 1.1 eq.).
The reaction mixture was stirred at r.t. for 20 h, and then the
mixture was concentrated in vacuo to give viscous oils, which were
purified by column chromatography (CH2Cl2/MeOH ¼ 10/1) to the
products (usually as yellow foams).
7.4) and the final reaction mixture was 200
mL. Antioxidant
(20 L) and fluorescein (120 L, 300 nM final concentration)
m
m
were placed in the wells of a 96 well plate and the mixture was
incubated for 15 min at 37 ꢃC. Then AAPH (Sigma, Steinheim
Germany) solution (60
mL; 12 mM final concentration) was
added rapidly. The plate was immediately placed into a Spec-
traFluor Plus plate reader (Tecan, Crailsheim, Germany) and
fluorescence measured every 60 s for 90 min with excitation at
485 nm and emission at 535 nm. 6-Hydroxy-2,5,7,8-
tetramethylchroman-2-carboxylic acid (Trolox, Sigma, Stein-
4.5.1. 14-Methyl-5,7,8,13,13b,14-hexahydroindolo[20,30:3,4]pyrido
[2,1-b]quinazolin-3-yl heptyl carbamate (11c)
Yellow foam (85 mg, yield 95%). 1H NMR (400 MHz, CDCl3)
d 8.31
(s, 1H, N13eH), 7.54 (d, J ¼ 7.8 Hz, 1H), 7.36 (d, J ¼ 8.0 Hz, 1H), 7.23e
7.17 (m, 1H), 7.13 (td, J ¼ 7.5, 1.0 Hz, 1H), 7.00 (d, J ¼ 8.7 Hz, 1H), 6.94
heim, Germany) was used as standard (1e8
mM, final

G. Huang et al. / European Journal of Medicinal Chemistry 81 (2014) 15e21
21
concentration). A blank (FL þ AAPH) using phosphate buffer
Acknowledgments
instead of antioxidant and Trolox calibration were carried out in
each assay. The samples were measured at different concen-
M. Decker gratefully acknowledges the German Science
Foundation (“Deutsche Forschungsgemeinschaft”) for financial
support (DFG DE1546/6-1). Appreciation is expressed to Professor
S. Elz at the University of Regensburg for providing his expertise
and facilities for HRMS and kinetic studies. We thank Gabi Brun-
ner for her excellent technical assistance on performing the ORAC
assay.
trations (1e5 mM). All reaction mixtures were prepared fourfold
and at least four independent runs were performed for each
sample. Fluorescence measurements were normalized to the
curve of the blank (without antioxidant). From the normalized
curves, the area under the fluorescence decay curve (AUC) was
calculated as:
AUC ¼ 1 þ i ¼ 90 fi=f0
(1)
X
Appendix A. Supplementary data
i ¼ 1
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2014.05.002.
Where f0 is the initial fluorescence at 0 min and fi is the fluores-
cence at time i.
The net AUC for a sample was calculated as follows:
References
Net AUC ¼ AUCantioxidant ꢁ AUCblank
(2)
[1] Y. Huang, L. Mucke, Cell 148 (2012) 1204e1222.
[2] C. Ballard, S. Gauthier, A. Corbett, C. Brayne, D. Aarsland, E. Jones, Lancet 377
(2011) 1019e1031.
The ORAC-FL values were calculated:
[3] A. Takeda, E. Loveman, A. Clegg, J. Kirby, J. Picot, E.C. Payne, Int. J. Geriatr.
Psych. 21 (2006) 17e28.
[4] M. Singh, M. Kaur, H. Kukreja, R. Chugh, O. Silakari, D. Singh, Eur. J. Med. Chem.
70 (2013) 165e168.
½ðAUC Sample ꢁ AUC blankÞ = ðAUC Trolox ꢁ AUC blankÞꢄꢂ
½ðconcentration of Trolox=concentration of sampleÞꢄ
[5] X. Chen, M. Decker, Curr. Med. Chem. 20 (2013) 1673e1685.
[6] S. Darvesh, D.A. Hopkins, C. Geula, Nat. Rev. Neurosci. 4 (2003) 131e138.
[7] E. Giacobini, Pharmacol. Res. 50 (2004) 433e440.
[8] Y. Furukawa-Hibi, T. Alkam, A. Nitta, A. Matsuyama, H. Mizoguchi, K. Suzuki,
S. Moussaoui, Q.S. Yu, N.H. Greig, T. Nagai, K. Yamada, Behav. Brain. Res. 225
(2011) 222e229.
[9] N.H. Greig, T. Utsuki, D.K. Ingram, Y. Wang, G. Pepeu, C. Scali, Q.S. Yu,
J. Mamczarz, H.W. Holloway, T. Giordano, D. Chen, K. Furukawa,
K. Sambamurti, A. Brossi, D.K. Lahiri, Proc. Natl. Acad. Sci. U. S. A. 102 (2005)
17213e17218.
(3)
and expressed as Trolox equivalents by using the standard curve
calculated for each assay. Final results were in M of Trolox
equivalent/ M of compound.
m
m
4.8. Neuroprotection and neurotoxicity assay
[10] P.W. Elsinghorst, C.M. Tanarro, M. Gütschow, J. Med. Chem. 49 (2006) 7540e
7544.
[11] S. Rizzo, C. Rivière, L. Piazzi, A. Bisi, S. Gobbi, M. Bartolini, V. Andrisano,
F. Morroni, A. Tarozzi, J.P. Monti, A. Rampa, J. Med. Chem. 51 (2008) 2883e
2886.
[12] C.G. Carolan, G.P. Dillon, D. Khan, S.A. Ryder, J.M. Gaynor, S. Reidy, J.F. Marquez,
M. Jones, V. Holland, J.F. Gilmer, J. Med. Chem. 53 (2010) 1190e1199.
[13] E. Rivera-Becerril, P. Joseph-Nathan, V.M. Pérez-Alvarez, M.S. Morales-Ríos,
J. Med. Chem. 51 (2008) 5271e5284.
[14] Q. Yu, H.W. Holloway, T. Utsuki, A. Brossi, N.H. Greig, J. Med. Chem. 42 (1999)
1855e1861.
[15] D. Karlsson, A. Fallarero, G. Brunhofer, P. Guzik, M. Prinz, U. Holzgrabe,
T. Erker, P. Vuorela, Eur. J. Pharm. Sci. 45 (2012) 169e183.
[16] M. Decker, F. Krauth, J. Lehmann, Bioorg. Med. Chem. 14 (2006) 1966e1977.
[17] X. Chen, I.G. Tikhonova, M. Decker, Bioorg. Med. Chem. 19 (2011) 1222e1235.
[18] M. Decker, J. Med. Chem. 49 (2006) 5411e5413.
[19] F.H. Darras, B. Kling, J. Heilmann, M. Decker, ACS Med. Chem. Lett. 3 (2012)
914e919.
[20] A.M. Palmer, Trends Pharmacol. Sci. 32 (2011) 141e147.
[21] S.H. Lee, J.K. Son, B. Se Jeong, T.C. Jeong, H.W. Chang, E.S. Lee, Y. Jahng, Mol-
ecules 13 (2008) 272e300.
Cells and cell culture: HT-22 cells [29,30] were derived from
murine hippocampal tissue [35] and were kindly provided by the
Max Planck Institute of Psychiatry, Munich. HT-22 cells are grown
in high glucose Dulbecco’s Modified Eagle’s Medium (DMEM,
Invitrogen, Karlsruhe, Germany) supplemented with 10% (v/v)
heat-inactivated fetal calf serum (FCS) (Biochrom, Berlin, Germany).
Cells were kept under standard cell culture conditions at 37 ꢃC
under 5% CO2 in a humidified incubator. Cells were subcultured
every 2 days.
Cell viability was determined in a 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyl tetrazolium bromide (MTT) assay [36] under recently
described conditions [19,31,32]. Briefly, cells were seeded in 96-
well plates at a density of 5 ꢂ 103 per well and cultured for 24 h.
Subsequently cells were incubated for another 24 h either with
medium, compounds, or solvent only in presence (neuroprotection
assay) or absence (neurotoxicity assay) of 5 mM glutamate (mon-
[22] L. Wang, C.P. Hu, P.Y. Deng, S.S. Shen, H.Q. Zhu, J.S. Ding, G.S. Tan, Y.J. Li, Planta
Med. 71 (2005) 416e419.
osodium-L-glutamate, Merck, Darmstadt, Germany). Quercetin
(Sigma, Steinheim, Germany) in a concentration of 25
m
M served as
[23] G. Dong, S. Wang, Z. Miao, J. Yao, Y. Zhang, Z. Guo, W. Zhang, C. Sheng, J. Med.
Chem. 55 (2012) 7593e7613.
[24] J. Jiang, C. Hu, Molecules 14 (2009) 1852e1859.
[25] W.F. Chiou, C.F. Chen, Eur. J. Pharmacol. 446 (2002) 151e159.
[26] S.M. Yuan, K. Gao, D.M. Wang, X.Z. Quan, J.N. Liu, C.M. Ma, C. Qin, L.F. Zhang,
Acta Pharm. Sin. 32 (2011) 295e302.
positive control in the neuroprotection assay. MTT (Sigma, Stein-
heim, Germany) solution (4 mg/mL in PBS) was diluted 1:10 with
medium and the mixture was added to the wells after removal of
previous medium. The plates were then incubated for another 3 h.
[27] J.Y. Lee, M.R. Cha, C.W. Choi, Y.S. Kim, B.H. Lee, S.Y. Ryu, Korean J. Pharmacogn.
43 (2012) 122e126.
Afterward, supernatants were removed and 100 mL of lysis buffer
(10% SDS) was added to the wells. Absorbance at 560 nM was
determined on the next day with a multiwell plate photometer
(SpectraFluor Plus, Crailsheim, Germany).
[28] G.L. Ellman, K.D. Coutney, V. Anders, R.M. Featherstone, Biochem. Pharmacol.
7 (1961) 88e95.
[29] J.B. Davis, P. Maher, Brain Res. 652 (1994) 169e173.
[30] S. Tan, D. Schubert, P. Maher, Curr. Top. Med. Chem. 1 (2001) 497e506.
[31] X. Chen, K. Zenger, A. Lupp, B. Kling, J. Heilmann, C. Fleck, B. Kraus, M. Decker,
J. Med. Chem. 55 (2012) 5231e5242.
[32] B. Kling, D. Bücher, P. Palatzky, F.M. Matysik, M. Decker, J. Wegener,
J. Heilmann, J. Nat. Prod. 77 (2014) 446e454.
[33] B. Ou, M. Hampsch-Woodill, R.J. Prior, J. Agric. Food. Chem. 49 (2001) 4619e
4926.
Results of cell viability are expressed as percentage to untreated
controlcells.AllcompoundsweredissolvedinDMSOanddilutedwith
fresh medium. DMSO concentration in final dilutions was ꢅ0.1%.
Statistical Analysis: Data are expressed as means ꢀ SD of at least
3 different independent experiments. Data were subjected to
oneway ANOVA followed by Dunnett’s multiple comparison post
test using GraphPad Prism 5 Software. (Levels of significance *
p < 0.05; **p < 0.01; ***p < 0.001).
[34] A. Dávalos, C. Gómez-Cordovés, B. Bartolomé, J. Agric. Food. Chem. 52 (2004)
48e54.
[35] B.H. Morimoto, D.E. Koshland, Neuron 5 (1990) 875e880.
[36] T. Mosmann, J. Immunol. Methods 65 (1983) 55e63.