Molecular Docking - Indolizine

Bioorganic Chemistry 116 (2021) 105390
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Interest of novel N-alkylpyridinium-indolizine hybrids in the field of

Alzheimer’s disease: Synthesis, characterization and evaluation of
antioxidant activity, cholinesterase inhibition, and amyloid
fibrillation interference
Isabelle Baussanne a, Olga Firstova a, b, Andreea Botezatu Dediu c, Camille Larosa b,
Bianca Furdui c, Ioana Ottilia Ghinea c, Aline Thomas a, Sabine Chierici b, *, Rodica Dinica c, *,
Martine Demeunynck a
a
Univ. Grenoble Alpes, CNRS, DPM, Grenoble, France
b
Univ. Grenoble Alpes, CNRS, DCM, Grenoble, France
c
Dunarea de Jos University of Galaţi, Faculty of Science and Environment, 111 Domneasca Street, 800201 Galaţi, Romania
A R T I C L E I N F O A B S T R A C T
Keywords: A small library of molecules combining indolizine and N-alkyl pyridinium was synthesized and evaluated in a
Indolizine-pyridinium multi-target-directed-ligand strategy for Alzheimer’s disease (AD) treatment. The new compounds were classified
Alzheimer’s disease in three series depending on the number of methylene residues linking the two heterocycles (Ind-PyCx with x =
Cholinesterase inhibition
0, 2 or 3). The molecules were synthesized from the corresponding bis-pyridines by two-step formation of the
Amyloid fibrillation
Antioxidant activity
indolizine core including mono-alkylation of pyridine and 1,3-dipolar cycloaddition with an alkylpropiolate.
Their activities against AD’s key-targets were evaluated in vitro: acetyl- and butyrylcholinesterase (AChE and
BChE) inhibition, antioxidant properties and inhibition of amyloid fibril formation. None of the three series
showed significant activities against all the targets. The Ind-PyC2 and Ind-PyC3 series are active on eeAChE and
hAChE (µM IC50 values). Most of the positively charged molecules from these two series also appeared active
against eqBChE, however they lost their activity on hBChE. Comparative molecular modeling of 13 and 15
docked in hAChE and hBChE highlighted the importance of the substituent (p-methoxybenzoyl or methylox
ycarbonyl, respectively) located on the indolizine C-3 for the binding. The larger molecule 13 fits more tightly at
the active site of the two enzymes than 15 that shows a larger degree of freedom. The Ind-PyC2 and Ind-PyC3
hybrids displayed some antioxidant activity when tested at 750 µg/mL (up to 95% inhibition of DPPH radical
scavenging for 10). In both series, most hybrids were also able to interact with amyloid fibers, even if the
inhibitory effect was observed at a high 100 µM concentration. The Ind-PyC0 molecules stand out completely
due to their spectroscopic properties which prevent their evaluation by Ellman’s and ThT assays. However, these
molecules showed interesting features in the presence of preformed fibers. In particular, the strong increase in
fluorescence of 3 in the presence of amyloid fibers is very promising for its use as a fibrillation fluorescent re
porter dye.
1. Introduction the clinic [2], however these drugs only delay the AD progression by
increasing the local acetylcholine level. In recent years the interest for
The Alzheimer disease (AD) is a multifactorial neurodegenerative these drugs has been questioned due to the existence of side-effects and
pathology characterized by oxidative stress and inflammatory processes, lack of proven long-term efficiency. Cholinesterase (ChE) inhibitors
deposition of senile plaques and neurofibrillary tangles, not to forget the targeting AChE and/or BChE (butyrylcholinesterase) remain the para
choline deficit due to neuron degeneration [1]. Up to now, acetylcho digm in the search of new drugs [3]. In an effort to improve the effec
linesterase (AChE) inhibitors are the only molecules that have entered tiveness of these treatments and considering the complexity of AD, the
* Corresponding authors.
E-mail addresses: [email protected] (S. Chierici), [email protected] (R. Dinica).
https://doi.org/10.1016/j.bioorg.2021.105390
Received 25 July 2021; Received in revised form 17 September 2021; Accepted 26 September 2021
Available online 2 October 2021
0045-2068/© 2021 Elsevier Inc. All rights reserved.
I. Baussanne et al. Bioorganic Chemistry 116 (2021) 105390
multi-target-directed ligands (MTDL) approach has gained a growing We describe here the synthesis of two other series that differ by the
importance [4–8]. Two or three AD-related features were simulta length of the aliphatic linker (2 or 3 methylenes for compounds 5–10
neously targeted by designing molecules that conjugate ChE inhibition and 11–16 respectively), and by the nature of their substituents
with other properties, such as antioxidant or β-amyloid anti-aggregation (Fig. 2A). This small library was evaluated against several current AD’s
activities. In a commonly used approach, heterodimers were prepared targets, i.e. AChE and BChE inhibition, antioxidant activity, and amyloid
by linking a known AChE inhibitor to a molecule aimed to a different fibril formation inhibition. The results were compared to those of the
activity or biological target [9,10]. A wide variety of such hybrids have corresponding bis-pyridinium salts 17–20 (Fig. 2B). Ellman and DPPH
been prepared as recently reviewed [11], including pyridinium con assays were used to respectively assess ChE inhibition and antioxidant
taining hybrids [12–19]. ability. In an effort to understand the factors involved in cholinesterase
Our groups have been involved for several years in the synthesis and inhibition, docking calculations were performed with the most relevant
study of the physical, chemical and biological properties of heterocyclic hybrid molecules, and the influence of the structures on the binding
compounds made of pyridinium salts and indolizines. Both heterocycles modes will be discussed. Finally, the interference of the molecules on
possess a wide variety of interesting properties [20–23] (for examples of amyloid fibrillation was investigated using short peptide models from
active molecules, see Fig. 1). It is worth highlighting the fluorescence tau protein [46–49].
properties of the indolizine nucleus which has been used in the design of
fluorescent tags or markers [24–30]. 2. Results and discussion
Interestingly in the field of AD, several pyridinium salts have been
designed to interfere with ChE activity as mentioned above 2.1. Syntheses
[9,10,14–16,31–35] and it is worth citing methoxime and obidoxime
that are potent agents against nerve toxin action [36]. Both pyridiniums The first objective was to prepare a series of N-alkylpyridinium-
and indolizines compounds also showed significant antioxidant prop indolizine hybrids with structural diversity by both playing with the
erties [37,38] or interference with amyloid aggregation [31,39,40]. nature of the starting bis-pyridines and with that of the N-alkyl sub
Starting from these considerations it was tempting to evaluate the stituents. Two strategies were envisioned to prepare the N-alkylpyr
indolizines and pyridinium derivatives that we are developing in our idinium-indolizine hybrids, both based on the formation of the
groups as MTDL’s against AD. Three series of derivatives were envi indolizine ring through 1,3-dipolar cycloaddition of alkynes with the
sioned as possible candidates: symmetrical bis-pyridinium salts or bis- ylides formed in situ from suitably substituted pyridinium salts. The
indolizines, and N-alkylpyridinium-indolizine hybrids. However, bis- synthetic pathways are depicted in Scheme 1.
pyridinium salts are already known structures in this field [15,41–43], As previously described, the Ind-PyC0 series, 2–4, were prepared
and the low solubility of bis-indolizines in water appeared as a severe from the non-symmetrical bis-alkylated bipyridinium salts, themselves
limitation to their biological evaluation. Nevertheless, preliminary in obtained by controlled mono-methylation of 4,4′ -bipyridine followed by
vestigations to this work confirmed the potential as antioxidant and a second alkylation with iodoacetophenones [44,45]. The cyclization
acetylcholinesterase inhibitors of both symmetrical bis-pyridiniums and step, with ethyl propiolate, regioselectively occurred on the ylide
bis-indolizines (unpublished data, see SI). To make the most of each forming side of the molecules (CH2COR).
heterocycle’s properties, we decided to focus on hybrid molecules con For the Ind-PyC2 and Ind-PyC3 series, with respectively 2 or 3
taining both an indolizine ring and a pyridinium moiety bearing various carbons between the two rings, the molecules mostly differ by the nature
N-substituents. of the pyridinium N-alkyl group. It then appeared more time saving to
A first series of N-alkylpyridinium-indolizines (Fig. 2A, Ind-PyC0 first build the indolizine core, and ultimately quaternize the resulting
series, compounds 1–4) was synthesized previously to this work [44,45]. pyridine-indolizine using various alkylating agents (Scheme 1). In most
Fig. 1. Examples of pyridinium salts and indolizine derivatives of biological interest.
2
Fig. 2. Hybrid molecules and unsymmetrical bis-pyridiniums studied in the present work.
cases, the first alkylation step was performed at 70–80 ◦ C in CH3CN in 2.2. ADMET molecular descriptors
the presence of 0.9 to 1 equivalent of the alkylating agent (2-bromo-p-
methoxyacetophenone or methyl 2-bromoacetate), to give the desired Considering the size and the positive charge carried by the hybrids, it
monoalkylated product as major compound (30–85% not optimized appeared important to check their ADMET descriptors. For a long time,
yields). In some cases, the mixture of mono- and bis-alkylated com uncharged small lipophilic molecules were the only candidates for
pounds thus obtained was directly engaged in the next step. To form the central nervous system (CNS) targeting, however the discovery of spe
indolizine ring, the reaction with methyl propiolate was achieved at cific Organic Cation Transporters (OCT), enabling charged molecules to
room temperature in CH2Cl2 using triethylamine as a base to generate pass through the blood brain barrier (BBB) to enter into the brain,
the reactive ylide. The resulting pyridine-indolizines 5,6,11,12 were changed this paradigm [50,51]. Actually, most CNS active drugs bear
isolated in 12–51% yields (two steps). The final alkylation step, with basic amines and are therefore positively charged at pH = 7.4 [52].
various alkylating agents, was done in CH3CN to give the pyridinium- To help efficient designing of new molecules as CNS candidates,
indolizine salts 7–10,13–16 in medium to excellent yields (50–100%). ADMET rules were adapted and a series of key parameters were selected
Four bis-pyridinium salts (17–20) were also prepared by controlled [52]: 1) Lipophilicity with the calculated partition coefficient LogP,
mono N-alkylation with methyl iodide, followed by the reaction with ideally found between 0.4 and 5.1; 2) Distribution at pH = 7.4 with LogD
bromoacetophenone (the structures are depicted in Fig. 2). ranging between 0.4 and 3.8; 3) Molecular weight, average = 305; 4)
Number of H-bond donors ≤ 2; and 5) Polarity with polar surface area
(PSA) ranging between 16 and 86 Å2. This last criterium has been
pointed out by other authors as being the most important factor, with
3
Scheme 1. General route to the N-alkylpyridinium-indolizine hybrids of the Ind-PyC2 and Ind-PyC3 series. The Ind-PyC0 compounds have been previously
described [44,45].
PSA < 90 [53], along with MW < 450 and LogD ranging between 0 and the more hydrophilic 3-hydroxypropyl group having the opposite effect.
3. In the C2 and C3 series, exchanging the more hydrophobic p-
We calculated a set of ADMET descriptors. The data are collected in methoxybenzoyl group on indolizine ring by the methyloxycarbonyl
Table S1. For comparison, we also calculated the data for the symmet group has no effect on PSA, but notably decreases LogP values (from
rical bis-pyridinium salts and bis-indolizines (1′a-c/2′a-c and 3′a-c/4′a- 0.65 for 13 to − 0.64 for 15, or from 2.06 for 14 to 0.77 for 16).
c respectively, Table S4). It is worth noting the discrepancy that may be Increasing the length of the linker, from 0 to 3 methylenes, has no effect
found using different calculation softwares, however our purpose was on PSA, but leads to an increase in lipophilicity (LogP ranging from
mainly the comparison of closely related structures. All calculations − 0.32 for 4 to 0.65 for 13).
were made using Marvin software. The BBB penetration ability was estimated using Swiss ADME web
The pyridine-indolizines (1, 5, 6, 11 and 12) are not or only partly service. The results predicted that the molecules showing high PSA and/
protonated at pH 7.4 (calculated pKa at 4.6 for 1 and 5.6 for the other or LogP (8, 9, 14, 16), i.e. containing large lipophilic substituents, are
compounds), leading to a large increase in lipophilicity reflected by not favorable for BBB crossing. However, the involvement of an active
higher LogP values. Alkylation of the pyridine ring gives mono-cationic penetration mechanism cannot be ruled out.
molecules in a large pH range, thus conferring to the molecules ADMET
values compatible with CNS potential drugs [54]. 2.3. Biological evaluation
In the Ind-PyC0 series, the major difference was found with 3, in
which the presence of a nitro group significantly increases the PSA value 2.3.1. Spectroscopic properties
to 94.8 Å2, i.e. the highest value in this hybrid library, compared to 60.9 The commonly used methods to evaluate cholinesterase inhibition
Å2 for the methoxy substituted analog 4. and antioxidant activity are Ellman’s and DPPH assays, respectively.
In the Ind-PyC2 series (compounds 7–9), the nature of the N-pyri They are both colorimetric spectroscopic assays. The Ellman’s method
dine substituent (methyl, 3,5-dimethoxybenzyl, 3-hydroxypropyl) in [55] needs 5,5′ -dithio-bis-2-nitrobenzoic acid (DTNB) and acetylth
fluences the PSA values in some extent (from 60.9 to 81.1). Replacing iocholine as alternative substrate of the enzyme. The AChE activity, in
methyl by the bulky 3,5-dimethoxybenzyl group increases LogP value, presence or absence of the tested compounds, is monitored by the
4
formation of the yellow 5-thio-2-nitrobenzoate that absorbs at 405 nm. standard deviation due to precipitation issue and are not therefore
For the evaluation of antioxidant ability, the assay is based on the indicated in Fig. 3.
reduction of a purple colored DPPH (2,2-diphenyl-1-picrylhydrazyl) Most of the molecules showed no or low activity when dosed at 48
that strongly absorbs at 517 nm [56]. DPPH is a free radical that can μg/ml, except with the bis-pyridinium salt 20, significantly active at that
react with a hydrogen donating antioxidant molecule. The antioxidant concentration (around 50% inhibition). However, rising the drug con
effect is then evaluated by following the decrease of the 517 nm ab centration up to 750 μg/ml led to a significant increase in efficiency for
sorption. Thus, if the molecules to be tested absorb strongly around most compounds, and in particular the bis-pyridinium 17–19 showed up
either 400 or 500 nm, results of assays may be biased. Absorption of the to >90% inhibition. Among the hybrids, the neutral compound 12
tested molecules is also of importance when studying their effect onto seemed promising with near 60% inhibition at 188 μg/ml, but solubility
amyloid fibril formation by the fluorescence Thioflavin (ThT) assay issues prevented correct measurements at higher concentration. Com
[57,58]. ThT assay is based on the fluorescence of the ThT dye in pound 10 showed the highest antioxidant properties (about 90% inhi
presence of the characteristic β-sheet motif of amyloid fibril in forma bition) at 750 μg/ml. Note that we observed similar trends with the
tion. The use of blank controls makes it possible to assess the intrinsic previously described symmetrical analogs: bis-pyridinium salts showed
fluorescence of compounds which could lead to erroneous in better activity compared to bis-indolizines (compare 1′ and 2′ series vs
terpretations. However, ThT assay may also be biased by the inner filter 3′ and 4′ series, Table S4).
effect of compounds with strong absorption at 440 nm or/and 480 nm,
corresponding to excitation and emission wavelengths of the bound ThT 2.3.3. Cholinesterase inhibition
[59,60].
In light of this, we first recorded the UV/Vis absorption of all the 2.3.3.1. IC50 determination. The cholinesterase activity was evaluated
molecules to check if their own spectroscopic properties would interfere using the Ellman’s spectrophotometric method [55] against electric eel
with the assays (Fig. S1). Except for the uncharged hybrid 1, the Ind- AChE (eeAChE) and equine butyrylcholinesterase (eqBChE), and
PyC0 molecules strongly absorb around 415 nm due to their extended selected compounds have been further tested against the human en
conjugation, as observed previously [44], making impossible their zymes (hAChE and hBChE).
evaluation with both Ellman’s and ThT assays. For all other compounds, Our goal being to select molecules that would significantly inhibit
maximum UV/Vis absorptions around 325 or 375 nm are suitable for the cholinesterases, a rapid screening of all series was first performed at 4
three assays. μM against eeAChE and eqBChE (see Figs. S2–S5). Only the molecules
that inhibited at least 50% of the ChE activity at this micromolar con
2.3.2. Antioxidant properties centration were then evaluated in a larger concentration range to pre
Oxidative stress is involved in age related neurodegenerative dis cisely calculate IC50 values. The data are collected in Table 1 (see also
eases and in particular in AD. Thus, antioxidants may have positive Table S3 for the IC50 curves).
benefits in reducing or delaying neuronal death. The antioxidant prop The already mentioned strong UV–Vis absorbance of the Ind-PyC0
erties of the molecules were evaluated by DPPH radical scavenging. The molecules, along with their lower solubility in water, prevented their
1,1-diphenyl-2-picrylhydrazyl or DPPH is a stable free radical charac study in a large range of concentrations, and the promising effects ob
terized by a deep violet color, with an absorption band in ethanol so tained in the 4 μM screenings, especially for 2 and 3 against eeAChE,
lution at 517 nm. In the presence of a molecule that can donate a could not be confirmed (see Figures S3). The bis-pyridinium salts 17–20
hydrogen atom, the formation of reduced DPPH is accompanied with the did not show any significant inhibition properties against the two en
loss of this violet color. Rutin was used as reference. The DPPH inhibi zymes at 4 µM (Fig. S5). Similarly, the uncharged pyridine-indolizines
tion % which relates to antioxidant activity was calculated at 48 μg/ml (6, 11 and 12) are inactive against both enzymes (Fig. S4). As ex
drug concentrations (i.e. the concentration for which rutin gives 83.3 ± pected, most of the corresponding symmetrical cationic pyridine-
0.1 inhibition% in our assay), 188 and 750 μg/ml drug concentrations indolizine hybrids were active at 4 µM. Unlike what was previously
(Fig. 3) [61]. All the molecules have been tested, however as indicated in observed with most of bis-indolizines (see Table S4) that easily form
Table S2, some molecules (compounds 1 and 3 from the Ind-PyC0 series aggregates in aqueous solutions, the generally better solubility of the
and the uncharged compounds 5 and 11) displayed an unacceptable pyridinium-indolizine hybrids allowed their evaluation in a larger
Fig. 3. Antioxidant activity (%) is measured using DPPH assay at 48 μg/ml (green), 188 μg/ml (blue) or 750 μg/ml (grey) drug concentrations, rutin was used as
reference. All assays were performed in triplicate and each bar value corresponds to the mean ± SD. Both the values with unacceptable SD and closed to zero are not
represented (see Table S2). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
5
Table 1 the tested enzymes, we wondered if these esters could be substrates of

Cholinesterase inhibition.a,b these enzymes [63]. To answer this question, the behavior of 15, con
Ind- 6 7 8 10 9 taining two methyl esters groups at positions 1 and 3 of the indolizine
PyC2 ring, was studied in presence of either hAChE or hBChE and of their
ee Inact.c 4.4 ± Inact.c 5.4 ± 2.0 ± 0.4 respective substrates (acetyl- or butyrylthiocholines). The reactions
AChE 0.5 0.1 were monitored by HPLC using a diode array detector, and after one day
eqBChE Inact.c
17.5 Inact. c
55.4 8.8 ± 3.5 of reaction the organic compounds were extracted from the pH8 buffer
± 8.2 ± 2.3 by dichloromethane. After evaporation of the organic phases, the resi
hAChE nd 6.6 ± nd 37.1 8.0 ± 0.9
1.1 ± 4.1
dues were dissolved in deuterated chloroform to perform 1H NMR
hBChE nd nd nd nd nd analysis. Neither retention time nor absorbance values appeared modi
fied on the HPLC chromatograms. The NMR data corresponded to a
Ind- 11 12 13 14 15 16 Donepezile
PyC3 mixture of the starting diester 15 and of thiocholine (Fig. S2). However,
the integration of the peaks at 3.95 ppm corresponding to the two OCH3
eeAChE Inact.c Inact.c 2.6 ± 7.9 ± 2.7 0.052 ±
appeared slightly smaller than expected for the diester, which may
>10
0.6 µMd 0.6 ± 0.007
0.6 indicate the presence of a small amount of partly hydrolyzed 15. The
eqBChE Inact.c Inact.c 7.5 ± 4.8 ± 72.1 7.3 6.0 ± 1.7 close proximity of the two OCH3 peaks did not allow the identification of
3.4 1.3 ± 15.9 ± the ester that was preferentially hydrolyzed. The dichloromethane
1.2
extraction from the pH8 buffer used to perform the enzymatic reaction
hAChE nd nd 3.3 ± nd 4.0 ± 9.4 0.029 ±
0.7 0.9 ± 0.005 constituted a limitation to this study as the hydrolysis of 15 would give
2.5 either a zwitterion (mono-hydrolysis) or a negatively charged diacid
hBChE nd nd 38.9 nd >100 nd 9.2 ± 5.6 (bis-hydrolysis), that would be difficult to extract from the basic
± µM medium.
14.0
However, this study clearly showed that after one day of reaction, a
a
IC50 values in µM are expressed as mean ± standard error of three inde significant amount of diester 15 was still present in solution, but a
pendent experiments performed in triplicates. partial hydrolysis could not be ruled out from the NMR analysis.
b
For easier comparison of C2 versus C3 series, the molecules containing the
same substituents are represented in the same column.
c 2.3.3.2. Docking calculations. Docking calculations were then per
The molecules indicated as inactive did not show significant activities during
formed to compare the possible binding modes of compounds 13 and 15
a first rapid screening at 4 μM (Figs. S3–S6).
d
Formation of aggregates at higher concentrations in presence of eeAChE. within the active sites of the two human ChE. Indeed, whilst both are
e
Donepezil is used as control (see Table S3 and Fig. S7 for IC50 curves). active as inhibitors of AChE (eel or human), they showed a different
behavior on BChE (equine or human).
concentration range. It is challenging to design a good inhibitor that would be active on
Thus, the cationic hybrids displayed micromolar IC50, with a slight both hAChE and hBChE, as these two enzymes differ considerably in size
selectivity for eeAChE vs eqBChE (Table 1), except 10 and its analog 15, (hAChE binding site is much smaller than hBChE) and in composition
which showed the highest selectivity (being 10 times more active (residues W286, V294, Y124, Y337, F297 in hAChE are respectively
against eeAChE compare to eqBChE) [62]. A277, P285, Q119, A328 and V288 in hBChE).
The Ind-PyC3 molecules containing the 3-p-methoxybenzoyl group Compounds 13 and 15 showed Glide docking scores of respectively
appeared more active against the two enzymes than their Ind-PyC2 − 11.3 and − 9.3 kcal/mol in hAChE, and − 9.1 kcal/mol and − 8.5 kcal/
analogs (compare 13 and 14 with 7 and 8, respectively). This effect may mol in hBChE. These theoretical affinity patterns are consistent with
be related to the better flexibility of the propyl linker compared to the experimental IC50, 13 being more active than 15 on both enzymes
shorter ethyl one. Concerning the eeAChE inhibition, the pyridinium- (Table 1).
indolizine hybrid 8 was the only inactive hybrid. This hybrid bears the In the hAChE active site (Fig. 4), the best pose of 13 had its pyr
lipophilic bulky N-dimethoxybenzyl group along with the 3-p-methox idinium moiety oriented towards the bottom of the catalytic site, in the
ybenzoyl substituent. Its Ind-PyC3 analog 14 displayed more interesting immediate vicinity of E202 and of the catalytic serine S203 (Fig. 4b).
activities against both enzymes, however a closer look to the curves and This binding mode was fully conserved among the 20 best poses of
Prism calculations (Table S3) showed that its limited solubility in 13, probably driven by the stabilizing interaction with E202 negative
presence of eeAChE at concentrations higher than 10 µM leads to a charge. Thus, 13 is not in a suitable position for hydrolysis by the hAChE
decreased R-squared value compared to the other tested molecules. enzyme. Compound 15 with a small methyl ester group was oriented
Compound 16, analog of 14 in which the 3-p-methoxybenzoyl was inversely compared to 13: the pyridinium group towards the entrance of
replaced by the smaller size methyloxycarbonyl group, exhibited a good the cavity, and the methyl ester group at position 3 facing the catalytic
activity against both eeAChE (IC50 = 2.7 μM) and eqBChE (IC50 = 7.3 triad at the bottom, allowing a possible hydrolysis (Fig. 4a). The indo
μM) without any apparent solubility limitation. lizine moiety is surrounded by residues W86, Y337, F338 and the pyr
Due to the high costs of the human cholinesterases, only a few active idinium moiety is in π -stacking interactions with W286. Due to its
molecules were then selected for further evaluation (Table 1, and Fig. S7 smaller size, 15 showed more heterogeneity among the binding poses.
for IC50 curves). For most tested compounds, IC50 against hAChE One of them, with the pyridinium moiety oriented towards the catalytic
remained in the same micromolar range, except for 10 that lost activity pocket and a salt bridge with E202 showed a close docking score of − 8.8
with IC50 > 30 μM for hAChE. Two molecules, 13 and 15, that only kcal/mol. These data are consistent with the similar activity observed
differed by the nature of the substituent at position 3 of the indolizine for both 13 and 15 in hAChE.
ring, were also tested against hBChE. The efficiency of 13, that previ In the hBChE binding site, 13 preferentially found its position with
ously showed a significant micromolar IC50 value against eqBChE, its pyridinium moiety inserted at the bottom of the site, close to the
strongly decreased (IC50 / 5) against hBChE. Compound 15 was not catalytic H438; the charge is stabilized by the hydroxyles of Y332 and
active (IC50 > 100 µM) confirming its previously observed selectivity for Y440 tyrosines, and by three carbonyl groups of residues A328 G78
AChE vs BChE. H438 (Fig. 4d). Residue W82 provides a favorable π-stacking interaction
The indolizine derived molecules reported in this study contain at to the pyridinium group. The methyl ester moiety is located at less than
least one carboxylic ester function. Considering the hydrolase activity of 5 Å to the catalytic S198, and we hypothesize that this compound may
6
Fig. 4. In a and b are illustrated the binding site of hAChE with respectively docked 15 and 13. In c and d are respectively drawn the best docked poses of 15 and 13
in the hBChE binding pocket. The catalytic residues are in yellow. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
adopt a closely-related binding mode to the natural substrate butyr displays a red shifted emission allowing the selective detection of β-sheet
ylcholine. This binding mode was very conserved among the twenty best rich structures such as fibers. We used here two models of amyloid
poses of 13. The best position of 15 showed its pyridinium moiety ori peptides, both derived from the tau protein sequence (Fig. 5).
ented towards the bottom of the cavity, where it is stabilized by the The AcPHF6 model is a short 6 amino-acid length peptide corre
carbonyl oxygen of A328 (Fig. 4c). The methyl ester at position 1 is in a sponding to the core of tau fibrils [49]. Its fibrillation is much faster than
perfect position for hydrolysis by S198. those of tau protein or amyloid peptides that can take days, and it is well
Compounds 13 and 15 showed a conserved orientation of their adapted for a first screening of a set of molecules by ThT fluorescence
pyridinium moieties within the hBChE pocket, but a reverse positioning assays. We already set up in previous works such screening in 96-well
of this group in hAChE: 13 has its pyridinium group towards the CAS microplates to study a series of aurones [46]. In addition, we used a
site, whereas 15 has its pyridinium group towards the entrance of the longer peptide derived from the R3 repeat tau region, AcR3, which has
cavity. Residue W286, which is an alanine in hAChE, is in pi-stacking been developed in our group (unpublished data). The studies were
interaction with the benzoyl substituent of 13 and the pyridinium of 15. performed in phosphate buffer, and the ThT fluorescence signal was
Concerning the diester 15, it is worth noting that with both enzymes, recorded in absence and in presence of the Ind-PyCx and BisPy mole
one methyl ester is in good position for hydrolysis, in agreement with the cules at a 100 µM concentration (corresponding to a 1/1 ratio with the
NMR study analysis. amyloid peptide). With AcR3 model, heparin was added as fibrillation
inducer, and 10 µM of heparin corresponding to 0.1/1 ratio with this
2.3.4. Amyloid fibrillation inhibition amyloid peptide was used. Results reported in Fig. 6 are expressed as the
To monitor the effect of molecules on the amyloid fibrillation pro percentage of ThT fluorescence inhibition. Blank wells were used to
cess, thioflavin-based fluorescence assay is generally used as it is highly heed any intrinsic fluorescence emission of the hybrids. No additional
sensitive to the β-sheet motifs of fibers. Upon binding to fibers, ThT fluorescence to that of free ThT was detected in these controls.
7
Fig. 5. Tau fibrillation process (in red β-sheet containing species), and sequences of the two tau models used in this study. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 6. ThT assays on AcPHF6 (A) and AcR3 (B) models of amyloid fibers (λexc = 440 nm, λem = 480 nm). The results are expressed as the inhibition % of ThT
fluorescence using a ratio compound/model of 1/1 (100 µM concentration). The errors represent the SEM of three independent experiments performed in triplicates.
As shown in Fig. 6A, most of the Ind-PyC2 and Ind-PyC3 hybrids PyC3 series. At first glance, the nature of the substituents carried by
displayed some activities on AcPHF6 model. The BisPy references each hybrid did not seem to be important either.
17–20 or the uncharged compound 6 and 12 were significantly less To go further, we decided to evaluate on the longer peptide model
efficient. The length of the Cx linker did not seem to play an important AcR3 the four most active Ind-PyC2 compounds (7–10) as well as the
role as similar activities were observed with both Ind-PyC2 and Ind- Ind-PyC0 1. The inhibition efficiency was then divided at least by a
Fig. 7. (A) ThT displacement assay. The excitation wavelength is 440 nm. The ratio compound/ThT is 10/1 (100/10 µM) and the AcR3 fibers have been prepared
from 100 µM of peptide in absence (solid lines) or in presence of ThT (dotted lines). (B) Fluorescence emission of Ind-PyC2 compounds (100 µM) alone (dotted lines)
or in presence of 100 µM of pre-formed fibrils (solid lines). The excitation wavelength corresponds to their absorbance maximum (value in parentheses); ThT in
presence of fibers is used for comparison.
8
factor two (compare Fig. 6A and 6B). Then, we recorded their fluorescence in presence of preformed fibers
It is noteworthy that inhibition of fluorescence in ThT assays does not as described above for the Ind-PyC2 series. With the most fluorescent
necessarily mean inhibition of amyloid fiber formation. Indeed, compounds 2 and 4, we observed an increase of their fluorescence
displacement of the bound ThT by the tested compounds may also occur. maximum with a slight red shift in presence of AcPHF6 fibers (Fig. 8B).
Consequently, we prepared a solution of AcR3 fibers in the presence of Such modification in the emission spectra may be explain by a close
ThT and recorded the bound ThT signal before and right after adding the interaction with fibers.
compounds of the Ind-PyC2 series (Fig. 7A, dotted lines). No decrease of More interestingly, among the non-fluorescent compounds 1 and 3,
ThT signal at 480 nm was observed suggesting that compounds are not only 3 showed a fluorescence light-up in presence of AcPHF6 fibers
able to displace the bound dye. The same observation was made with (Fig. 9A). In addition, the fluorescence intensity is much higher
compound 1. To ensure that the recorded signal was still from bound compared than that of the bound ThT dye at the same concentration.
ThT, we checked that the compounds fluorescence did not significantly The same response was observed, to a lesser extent, in the case of
light up in presence of preformed fibers at the working wavelengths AcR3 fibers (Fig. 9B and Fig. S10). Further study would be needed, but 3
(Fig. 7A, solid lines). However, the slight increase and red shift of the could be of interest as reporter dye of fibrillation. The fluorescence of 3
fluorescence signal observed in Fig. 7A (dotted lines) upon addition of being red-shifted in comparison to ThT (550 nm vs 480 nm), it would
the charged molecules 7–9 is probably due to this additional fluores allow screening of a wider range of potential inhibitors
cence, even weak, and suggests that 7–9 interact closely with fibers. This
is confirmed by their huge fluorescence emission in presence of fibers 3. Conclusion
when excited at their absorbance maximum while they are no fluores
cent alone (Fig. 7B). Compound 10, which bears two ester groups on its In the course of this study, we have synthesized and evaluated a
indolizine part (7 being its 3-p-methoxybenzoyl analog), is the exception series of 20 molecules, bis-pyridinium salts and pyridinium-indolizine
to the rule. Interestingly, analogs 14 and 16 from the Ind-PyC3 series hybrids. The molecules were tested against AD hallmarks, i.e. antioxi
exhibited the same behavior (Fig. S8): 16 with two ester groups did not dant properties, inhibition of AChE and BChE, and interference with
show a significative emission in presence of AcR3 fibers in contrast to amyloid fibrillation using tau peptide models. The hybrids are consti
14, suggesting the importance of the benzoyl substituent onto the tuted of an indolizine core linked to a pyridinium pendant, and they
indolizine part for a greater interaction with fibers. differ by the length of the linker, with 0, 2 or 3 methylene units (Ind-
Note that in the Ind-PyC0 series only the uncharged compound 1 PyC0, Ind-PyC2 and Ind-PyC3, respectively). The substituents at posi
was evaluated by ThT assay (Fig. 6). Indeed, as explained before, the tion 3 of the indolizine and on the heterocyclic pyridine nitrogen were
strong absorption of compounds 2–3 around 415 nm, and in addition the also varied. No hybrid showed significant activity on the three targets
strong fluorescence of 2 and 4 in the wavelength range of bound ThT simultaneously, i.e. antioxidant activity (DPPH inhibition %>50 at 48
fluorescence emission did not permit their evaluation [44] (Figs. S1 and µg/mL), hChE inhibition (IC50 < 10 μM) and inhibition of amyloid fiber
S9). formation. The Ind-PyC2 and Ind-PyC3 series were weak anti-oxidants,
Thus, to study the interaction of the Ind-PyC0 molecules, we used and moderate inhibitors of amyloid fibers formation. However, most of
the circular dichroism spectroscopy (CD) that provides structural in the positively charged molecules from these two series are active on
formation on amyloid peptides during the aggregation process. The CD both eeAchE and hChE at micromolar range. A few molecules also
spectra of AcR3 peptide were registered in phosphate buffer in presence inhibited eqBChE but unfortunately not the human enzyme hBChE. The
of heparin as fibrillation inducer. They displayed the typical profile of a Ind-PyC0 series showed a significantly different behavior from the other
peptide with a high fraction of β-sheet structures, characterized by a two series. Both solubility limitation (the molecules are prone to
broad negative shoulder around 218 nm (Fig. 8, red color) in less than aggregate in water solution) and spectral properties did not allow their
24 h at 37 ◦ C. No significative difference was noticed when compounds study neither as cholinesterase inhibitors nor as fibers inhibitors.
1–4 were added to the fibrillation mixture (Fig. 8A) attesting that Nevertheless, in the amyloid fiber context, fluorescence and CD exper
compounds of Ind-PyC0 series did not impede the formation of β-sheet iments revealed that if these molecules don’t interfere in the fibrillation
structures (note that the positive shoulder of β-sheet profile under 200 process, the positively charged 2–4, and in particular the nitro derived 3,
nm cannot be observed due to the strong absorption of DMSO used for may be new fluorescent reporter dyes of amyloid fiber formation.
the dilution of compounds). These results tend to show that 1–4 do not
inhibit fibers formation, or at least the formation of β-sheet-rich
structures.
Fig. 8. (A) CD spectra obtained after a 24 h incubation at 37 ◦ C of AcR3 at 100 µM, alone (red) or in presence of 100 µM of Ind-PyC0 series 1 to 4. (Green, blue, grey
and pink). 10 µM heparin is used as fibrillation inducer. (B) Fluorescence emission of Ind-PyC0 series (100 µM) alone (dotted lines) and in presence of 100 µM of pre-
formed AcPHF6 fibrils (solid lines). The excitation wavelength corresponds to their absorbance maximum (value in parentheses); 10 µM ThT in presence of fibers is
used for comparison (red solid line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
9
Fig. 9. Fluorescence emission of compounds 1 and 3 (100 µM) alone (dotted lines) and in presence (solid lines) of (A) 100 µM of pre-formed AcPHF6 fibrils or (B)
100 µM of pre-formed AcR3 fibrils (B). The excitation wavelength corresponds to their absorbance maximum (value in parentheses); 10 µM ThT in presence of fibers
is used for comparison.
4. Material and methods with water before being dried on MgSO4 and concentrated under
reduced pressure. The residue was then purified by flash chromatog
4.1. Synthesis raphy on silica gel (CH2Cl2/MeOH: 98/2) to obtain indolizine 5 in 61%
yield.
1
NMR spectra were recorded at room temperature in 5 mm tubes on a H NMR (400 MHz, CDCl3) δ 9.84 (dd, 1H, J = 7.2, 0.8 Hz), 8.56 (d,
Bruker AC 400 MHz spectrometer (NMR facility, PCN-ICMG, Grenoble). 2H, J = 4.4 Hz), 8.21 (d, 1H, J = 0.8 Hz), 7.87 (d, 2H, J = 8.8 Hz), 7.82
Chemical shifts (δ) are reported in parts per million (ppm) from low to (s, 1H), 7.41 (d, 2H, J = 6.0 Hz), 7.18 (d, 2H, J = 4.4 Hz), 7.04 (d, 2H, J
high field and referenced to residual non-deuterated solvent relative to = 8.8 Hz), 6.92 (dd, 1H, J = 7.2, 2.0 Hz), 3.94 (s, 3H), 3.93 (s, 3H), 3.09
Me4Si. Standard abbreviations for multiplicity were used as follows: s = (m, 4H); 13C NMR (100 MHz, CDCl3) δ 184.4, 164.6, 162.6, 149.9,
singlet; d = doublet; t = triplet; m = multiplet. High resolution mass 149.5, 141.3, 140.0, 132.3, 131.2, 128.9, 128.5, 123.9, 122.5, 117.7,
sectrometry (HRMS) was carried out on a Bruker UHR-Q-TOF MaXis- 116.4, 113.7, 104.9, 55.5, 51.2, 36.1, 35.7; HRMS (ESI) m/z: calcd. for
ETD (Time of Flight) mass spectrometer using ElectroSpray Ionisation C25H23N2O4 [M + H]+ 415.1652, obsd 415.1655.
(ESI) in Institut de Chimie Organique et Analytique (CBM-ICOA) in
Orleans (France). 4.1.1.2. Methyl 3-(p-methoxybenzoyl)-7-[2-(1-methylpyridinium-4-yl)
The N-alkylpyridinium-indolizines 1–4 of the Ind-PyC0 series ethyl]indolizine-1-carboxylate iodide (7). The indolizine 5 (372 mg, 0.9
[44,45] and the bis-pyridinium salt 17[64] have already been described. mmol) was dissolved in CH3CN (8 mL). Iodomethane (1.5 eq.) was
added and the solution was stirred at 50 ◦ C for 4 h. After concentration
4.1.1. Ind-PyC2 series under vacuum, the brown residue was washed with CH2Cl2. After dry
ing, the N-alkylpyridinium-indolizine 7 was thus isolated as a brown
4.1.1.1. Methyl 3-(p-methoxybenzoyl)-7-[2-(4-pyridinyl)ethyl] indolizine- solid with a quantitative yield.
1
1-carboxylate (5). This compound was prepared in two steps from 1,2- H NMR (400 MHz, CDCl3) δ 9.70 (d, 1H, J = 7.2 Hz), 9.04 (d, 2H, J
bis(4-pyridyl)ethane: = 6.4 Hz), 8.05 (s, 1H), 7.87 (d, 2H, J = 6.4 Hz), 7.74 (d, 2H, J = 8.4 Hz),
7.69 (s, 1H), 6.94 (s, 1H), 6.93 (d, 2H, J = 8.8, Hz), 4.57 (s, 3H), 3.83 (s,
(1) N-Alkylation 3H), 3.82 (s, 3H), 3.30 (t, 2H, J = 7.8 Hz), 3.13 (t, 2H, J = 7.8 Hz); 13C
NMR (100 MHz, CDCl3) δ 184.4, 164.6, 162.7, 161.0, 145.1, 139.6,
1,2-Bis(4-pyridyl)ethane (504 mg, 2.7 mmol) was dissolved in 139.2, 132.1, 131.2, 129.3, 128.5, 128.3, 122.7, 117.9, 116.3, 113.8,
CH3CN (7 mL). 2-Bromo-p-methoxyacetophenone (1 eq.) was added and 105.1, 55.5, 51.4, 49.1, 36.2, 35.0; HRMS (ESI) m/z: calcd. for
the solution was stirred at 70–80 ◦ C for 2 h. The yellow precipitate was C26H25N2O4 [M]+ 429.1809, obsd 429.1810.
filtrated and dried under vacuum to give pure 1-(2-(4-methoxyphenyl-2-
oxoethyl)-4-[2-(4-pyridinyl)ethyl]pyridinium bromide in 85% (945 4.1.1.3. Methyl 3-(p-methoxybenzoyl)-7-[2-(1-[(3,5-dimethoxyphenyl)
mg). methyl]pyridinium-4-yl)ethyl] indolizine-1-carboxylate bromide (8). The
1
H NMR (400 MHz, CD3OD) δ 8.77 (d, 2H, J = 6.4 Hz), 8.49 (d, 2H, J indolizine 5 (125.5 mg, 0.3 mmol) was dissolved in CH3CN (3 mL). 3,5-
= 6.0 Hz), 8.10 (m, 4H), 7.41 (d, 2H, J = 6.0 Hz), 7.16 (d, 2H, J = 8.8 Dimethoxybenzyl bromide (1.2 eq.) was added and the solution was
Hz), 5.53 (s, 2H), 3.97 (s, 3H), 3.44 (t, 2H, J = 7.8 Hz), 3.24 (t, 2H, J = stirred at 70–80 ◦ C for 3 h. After concentration under vacuum, the
7.8 Hz); 13C NMR (100 MHz, D2O) δ 190.8,165.0, 163.0, 150.5, 148.8, brown residue was washed with CH2Cl2. After drying, the residue was
145.1, 131.2, 128.0, 126.1, 124.7, 114.7, 55.9, 35.5, 34.1; HRMS (ESI) then purified by flash chromatography on silica gel (CH2Cl2/MeOH: 90/
m/z: calcd. for C21H21N2O2 [M]+ 333.1597, obsd 333.1595 and calcd. 10) to obtain the N-alkylpyridinium-indolizine 8 as a brown solid in 77%
for C21H22N2O2 [M + H]2+ 167.0835, obsd 167.0838. yield.
1
H NMR (400 MHz, CDCl3) δ 9.73 (d, 1H, J = 7.2 Hz), 9.30 (d, 2H, J
(2) Indolizine formation = 6.0 Hz), 8.07 (s, 1H), 7.80–7.42 (m, 4H), 7.70 (s, 1H), 6.94 (s, 1H),
6.93 (d, 2H, J = 8.8, Hz), 6.87 (dd, 1H, J = 7.2, 1.6 Hz), 6.73 (d, 2H, J =
The pyridinium salt thus obtained (600 mg, 1.45 mmol) and methyl 2.0 Hz), 6.37 (d, 1H, J = 2.0 Hz), 6.03 (s, 2H), 3.83 (s, 3H), 3.82 (s, 3H),
propiolate (2.2 eq.) were dissolved in anhydrous CH2Cl2 (10 mL). NEt3 3.71 (s, 6H), 3.23 (t, 2H, J = 7.6 Hz), 3.09 (t, 2H, J = 7.6 Hz); 13C NMR
(1.1 eq.) diluted in 5 mL of anhydrous CH2Cl2 was then added dropwise (100 MHz, CDCl3) δ.184.5, 164.6, 162.7, 161.7, 161.1, 144.5, 139.7,
in the reacting mixture. After 5 h stirring at rt, water was added to the 139.1, 134.6, 132.2, 131.2, 129.3, 128.4, 128.0, 122.7, 117.8, 116.0,
solution. The organic phase was separated and washed 2 more times 113.8, 107.5, 105.1, 102.0, 63.9, 55.9, 55.5, 51.3, 36.0, 34.8; HRMS
10
(ESI) m/z: calcd. for C34H33N2O6 [M]+ 565.2333, obsd 565.2337. chromatography on silica gel (CH2Cl2/MeOH: 96/4 to 90/10) to obtain
the N-alkylpyridinium-indolizine 10 was thus isolated as a yellow solid
4.1.1.4. Methyl 3-(p-methoxybenzoyl)-7-[2-(1-(3-hydroxypropyl)pyr in 85% yield.
1
idinium-4-yl)ethyl] indolizine-1-carboxylate bromide (9). The indolizine 5 H NMR (400 MHz, CDCl3) δ 9.49 (d, 1H, J = 6.8 Hz), 9.10 (d, 2H, J
(231.5 mg, 0.56 mmol) was dissolved in CH3CN (5 mL). 3-Bromo-1- = 5.6 Hz), 8.12 (s, 1H), 7.99 (s, 1H), 7.93 (d, 2H, J = 5.6 Hz), 6.92 (d,
propanol (3 eq.) was added and the solution was stirred at 70–80 ◦ C 1H, J = 6.8 Hz), 4.67 (s, 3H), 3.96 (s, 3H), 3.95 (s, 3H), 3.39 (t, 2H, J =
for 20 h. After concentration under vacuum, the residue was then pu 7.6 Hz), 3.22 (t, 2H, J = 7.8 Hz); 13C NMR (100 MHz, CDCl3) δ164.6,
rified by flash chromatography on silica gel (CH2Cl2/MeOH: 95/5 to 85/ 161.4, 161.1, 145.1, 138.9, 137.6, 128.3 128.0, 124.1, 117.9, 115.7,
15) to obtain the N-alkylpyridinium-indolizine 9 as a yellow solid in 114.5, 104.4, 51.5, 51.3, 49.0, 36.2, 34.9; HRMS (ESI) m/z: calcd. for
50% yield. C20H21N2O4 [M]+ 353.1496, obsd 353.1497.
1
H NMR (400 MHz, CD3OD) δ 9.81 (d, 1H, J = 7.2 Hz), 8.88 (d, 2H, J
= 6.4 Hz), 8.16 (s, 1H), 8.04 (d, 2H, J = 6.8 Hz), 7.85 (d, 2H, J = 8.8 Hz), 4.1.2. Ind-PyC3 series
7.78 (s, 1H), 7.20 (dd, 1H, J = 7.2, 1.6 Hz), 7.13 (d, 2H, J = 8.8, Hz),
4.72 (t, 2H, J = 7.0 Hz), 3.95 (s, 3H), 3.91 (s, 3H), 3.63 (t, 2H, J = 5.6 4.1.2.1. Methyl 3-(p-methoxybenzoyl)-7-[3-(4-pyridinyl)propyl]
Hz), 3.46 (t, 2H, J = 7.4 Hz), 3.32 (t, 2H, J = 7.2 Hz), 2.21 (m, 2H); 13C indolizine-1-carboxylate (11). This compound was prepared in two steps
NMR (100 MHz, DMSO‑d6) δ 183.3, 163.5, 162.2, 160.8, 144.3, 141.4, from 1,3-bis(4-pyridyl)propane
139.0, 134.6, 132.2, 131.5, 130.9, 128.4, 127.7, 127.3, 121.9, 117.2,
117.0, 113.9, 104.0, 57.9, 57.1, 55.5, 51.1, 34.9, 34.2, 33.2; HRMS (ESI) (1) N-Alkylation
m/z: calcd. for C28H29N2O5 [M]+ 473.2071, obsd 473.2074 and calcd.
for C28H30N2O5 [M + H]2+ 237.1072, obsd 237.1071. 1,3-Bis(4-pyridyl)propane (676 mg, 3.4 mmol) was dissolved in
CH3CN (8 mL). 2-Bromo-p-methoxyacetophenone (1 eq.) was added and
4.1.1.5. Dimethyl 7-[2-(4-pyridinyl)ethyl] indolizine-1,3-dicarboxylate the solution was stirred at 70–80 ◦ C for 2 h. The white precipitate was
(6). This compound was prepared in two steps from 1,2-bis(4-pyridyl) filtrated and then purified by flash chromatography on silica gel
ethane: (CH2Cl2/MeOH: 90/10). The 1-(2-(4-methoxyphenyl-2-oxoethyl)-4-[3-
(4-pyridinyl)propyl]pyridinium bromide was thus isolated as a white
(1) N-Alkylation solid in 41% yield (600 mg).
1
H NMR (400 MHz, CD3OD) δ8.76 (d, 2H, J = 6.8 Hz), 8.49 (dd, 2H,
1,2-Bis(4-pyridyl)ethane (504 mg, 2.74 mmol) was dissolved in J = 4.8, 1.6 Hz), 8.13–8.08 (m, 4H), 7.40 (dd, 2H, J = 4.8, 1.6 Hz), 7.17
CH3CN (15 mL). Methyl bromoacetate (0.9 eq.) was added and the so (dd, 2H, J = 6.8, 2.0 Hz), 3.97 (s, 3H), 3.11 (t, 2H, J = 7.8 Hz), 2.87 (t,
lution was stirred at 70–80 ◦ C for 3 h. The white precipitate (dialkylated 2H, J = 7.8 Hz), 2.21 (m, 2H); 13C NMR (100 MHz, CD3OD) δ188.2,
product) was removed by filtration. After concentration under vacuum, 165.2, 163.7, 152.1, 148.6, 145.4, 130.5, 127.4, 127.3, 126.3, 124.3,
water was added then the aqueous solution was washed 4 times with 114.1, 54.9, 34.7, 34.0, 29.6; HRMS (ESI) m/z: calcd. for C22H23N2O2
AcOEt to remove the residual 1,2-bis(4-pyridyl)ethane. After aqueous [M]+ 347.1754, obsd 347.1754 and calcd. for C22H24N2O2 [M + H]2+
phase concentration under vacuum, the residue was washed twice with 174.0913, obsd 174.0917.
CH3CN. The organic solution was concentrated in vacuo to give an or
ange oil composed of a 4/1 mixture of 1-(2-methoxy-2-oxoethyl)-4-[2- (2) Indolizine formation
(4-pyridinyl)ethyl]pyridinium bromide and 4,4′ -(1,2-ethanediyl)bis[1-
(2-methoxy-2-oxoethyl)pyridinium] dibromide in 30% yield. The The pyridinium salt thus obtained (540 mg, 1.26 mmol) and methyl
mixture was directly used for the next step without purification. propiolate (2.2 eq.) were dissolved in anh. CH2Cl2 (10 mL). NEt3 (1.2
eq.) diluted in 5 mL of anh. CH2Cl2 was then added dropwise in the
(2) Indolizine formation reacting mixture. After 2 more hours of stirring at rt, water was added to
the solution. The organic phase was separated and washed 2 more times
The pyridinium mixture (145 mg, 0.43 mmol) and methyl propiolate with water and brine before being dried on MgSO4 and concentrated
(2.1 eq.) were dissolved in anh. CH2Cl2 (4 mL). NEt3 (1.2 eq.) diluted in under reduced pressure. The residue was then purified by flash chro
1 mL of anh. CH2Cl2 was then added dropwise in the reacting mixture. matography on silica gel (CH2Cl2/MeOH: 98/2) to obtain indolizine 11
After 2 h stirring at rt. water was added to the solution. The organic in 68% yield.
1
phase was separated and washed 2 more times with water and brine H NMR (400 MHz, CDCl3) δ 9.74 (d, 1H, J = 7.2 Hz), 8.44 (d, 2H, J
before being dried on MgSO4 and concentrated under reduced pressure. = 5.6 Hz), 8.10 (s, 1H), 7.75 (d, 2H, J = 8.8 Hz), 7.71 (s, 1H), 7.06 (d,
The residue was then purified by flash chromatography on silica gel 2H, J = 6.0 Hz), 6.93 (d, 2H, J = 8.8 Hz), 6.83 (dd, 1H, J = 7.2, 1.6 Hz),
(CH2Cl2/MeOH: 98/2) to obtain indolizine 6 as an amorphous orange 3.82 (s, 6H), 2.71 (t, 2H, J = 7.6 Hz), 2.62 (t, 2H, J = 7.6 Hz), 1.99 (m,
solid in 40% yield. 4H); 13C NMR (100 MHz, CDCl3) δ 184.4, 164.7, 162.6, 150.5, 149.8,
1 142.4, 140.2, 132.4, 131.2, 128.9, 128.6, 123.9, 122.5, 117.7, 116.5,
H NMR (400 MHz, CDCl3) δ 9.46 (dd, 1H, J = 7.2, 0.8 Hz), 8.56 (dd,
2H, J = 4.4, 1.6 Hz), 8.19 (d, 1H, J = 0.8 Hz), 7.99 (s, 1H), 7.17 (dd, 2H, 113.7, 104.7, 55.5, 51.2, 35.0, 34.6, 30.7; HRMS (ESI) m/z: calcd. for
J = 4.4, 1.6 Hz), 6.84 (dd, 1H, J = 7.2, 2.0 Hz), 3.96 (s, 3H), 3.95 (s, 3H), C26H25N2O4 [M + H]+ 429.1809, obsd 429.1810.
3.07 (m, 4H); 13C NMR (100 MHz, CDCl3) δ 164.7, 161.6, 150.0, 149.5,
139.5, 139.4, 127.8 124.5, 123.8, 117.9, 115.9, 114.2, 104.2, 51.4, 4.1.2.2. Methyl 3-(p-methoxybenzoyl)-7-[3-(1-methylpyridinium-4-yl)
51.2, 36.0, 35.8; HRMS (ESI) m/z: calcd. for C19H19N2O4 [M + H]+ propyl]indolizine-1-carboxylate iodide (13). The indolizine 11 (102 mg,
339.1339, obsd 339.1338. 0.24 mmol) was dissolved in CH3CN (4 mL). Iodomethane (1.5 eq.) was
added and the solution was stirred at 50 ◦ C for 4 h. After concentration
4.1.1.6. Dimethyl 7-[2-(1-methylpyridinium-4-yl)ethyl]indolizine-1,3- under vacuum, the brown residue was washed with CH2Cl2. After dry
dicarboxylate iodide (10). The indolizine 6 (42 mg, 0.12 mmol) was ing, the N-alkylpyridinium-indolizine 13 was thus isolated as an orange
dissolved in CH3CN (2 mL). Iodomethane (1.5 eq.) was added and the solid with a quantitative yield.
1
solution was stirred at 50 ◦ C for 2 h before adding another portion of H NMR (400 MHz, CDCl3) δ 9.71 (d, 1H, J = 7.2 Hz), 9.04 (d, 2H, J
iodomethane (1.5 eq.). After 2 h more stirring, the solution was = 6.0 Hz), 8.08 (s, 1H), 7.83 (d, 2H, J = 6.4 Hz), 7.75 (d, 2H, J = 8.4 Hz),
concentrated under vacuum. The residue was then purified by flash 7.70 (s, 1H), 6.93 (d, 2H, J = 8.4, Hz), 6.90 (dd, 1H, J = 7.2, 1.6 Hz),
11
4.57 (s, 3H), 3.83 (s, 3H), 3.82 (s, 3H), 2.91 (t, 2H, J = 7.6 Hz), 2.79 (t, dissolved in CH3CN (2 mL). Iodomethane (3 eq.) was added and the
2H, J = 7.6 Hz), 2.08 (m, 2H); 13C NMR (100 MHz, CDCl3) δ 184.4, solution was stirred at 60 ◦ C for 3 h. After concentration under vacuum,
164.6, 162.6, 162.3, 145.0, 141.1, 139.9, 132.2, 131.2, 129.1, 128.5, the brown residue was washed with CH2Cl2 and Et2O. After drying, the
128.1, 122.6, 117.8, 116.4, 113.8, 104.9, 55.5, 51.3, 49.0, 35.1, 34.7, N-alkylpyridinium-indolizine 15 was thus isolated as a brownish solid
30.0; HRMS (ESI) m/z: calcd. for C27H27N2O4 [M]+ 443.1965, obsd (71%).
1
443.1966. H NMR (400 MHz, CDCl3) δ 9.46 (d, 1H, J = 7.2 Hz), 9.15 (d, 2H, J
= 6.4 Hz), 8.15 (s, 1H), 7.98 (s, 1H), 7.93 (d, 2H, J = 6.4 Hz), 6.93 (d,
4.1.2.3. Methyl 3-(p-methoxybenzoyl)-7-[3-(1-[(3,5-dimethoxyphenyl) 1H, J = 7.2 Hz), 4.68 (s, 3H), 3.95 (s, 6H), 3.02 (t, 2H, J = 7.6 Hz), 2.87
methyl]pyridinium-4-yl)propyl] indolizine-1-carboxylate bromide (14). (t, 2H, J = 7.6 Hz), 2.18 (m, 2H); 13C NMR (100 MHz, CDCl3) δ 164.7,
The indolizine 11 (193 mg, 0.45 mmol) was dissolved in CH3CN (4 mL). 162.3, 161.5, 145.0, 139.3, 128.0 127.9, 124.5, 117.9, 115.9, 114.3,
3,5-Dimethoxybenzyl bromide (1.2 eq.) was added and the solution was 104.2, 51.5, 51.3, 49.0, 35.1, 34.6, 30.0; HRMS (ESI) m/z: calcd. for
stirred at 70–80 ◦ C for 3 h. After concentration under vacuum, the C21H23N2O4 [M]+ 367.1652, obsd 367.1651.
brown residue was washed with CH2Cl2. After drying, the N-alkylpyr
idinium-indolizine 14 was thus isolated as a brown solid with a quan 4.1.2.6. Dimethyl 7-[3-(1-[(3,5-dimethoxyphenyl)methyl]pyridinium-4-
titative yield. yl)propyl]indolizine1,3-dicarboxylate bromide (16). The indolizine 12
1
H NMR (400 MHz, CDCl3) δ 9.73 (d, 1H, J = 7.2 Hz), 9.31 (d, 2H, J (58 mg, 0.14 mmol) was dissolved in CH3CN (2 mL). 3,5-Dimethoxyben
= 6.4 Hz), 8.09 (s, 1H), 7.77–7.71 (m, 5H), 6.94 (dd, 2H, J = 6.8, 2.0 zyl bromide (1.1 eq.) was added and the solution was stirred at 70 ◦ C for
Hz), 6.85 (d, 1H, J = 7.2 Hz), 6.76 (d, 2H, J = 2.0 Hz), 6.36 (dd, 1H, J = 3 h. After concentration under vacuum, the brown residue was washed
2.0 Hz), 6.04 (s, 2H), 3.83 (s, 3H), 3.82 (s, 3H), 3.71 (s, 6H), 2.85 (t, 2H, with CH2Cl2 and Et2O. After drying, the N-alkylpyridinium-indolizine
J = 7.6 Hz), 2.77 (t, 2H, J = 7.6 Hz), 2.04 (m, 2H); 13C NMR (100 MHz, 16 was thus isolated as a brownish solid in 90% yield.
1
CDCl3) δ 184.4, 164.6, 162.6 162.2, 161.7, 144.4, 141.0, 139.9, 134.7, H NMR (400 MHz, CDCl3) δ 9.47 (d, 1H, J = 7.2 Hz), 9.43 (d, 2H, J
132.3, 131.2, 129.1, 128.5, 127.8, 122.6, 117.7, 116.3, 113.8, 107.5, = 5.6 Hz), 8.16 (s, 1H), 7.98 (s, 1H), 7.83 (d, 2H, J = 5.6 Hz), 6.88 (m,
104.9, 102.0, 63.8, 55.9, 55.5, 51.3, 35.1, 34.8, 29.8; HRMS (ESI) m/z: 3H), 6.48 (s, 1H), 6.15 (s, 2H), 3.96 (s, 6H), 3.83 (s, 6H), 2.95 (t, 2H, J =
calcd. for C35H35N2O6 [M]+ 579.2490, obsd 579.2495 and calcd. for 7.6 Hz), 2.85 (t, 2H, J = 7.6 Hz), 2.13 (m, 2H); 13C NMR (100 MHz,
C35H36N2O6 [M + H]2+ 290.1281, obsd 290.1279. CDCl3) δ 164.7, 162.3, 161.7, 161.5, 144.4, 139.3, 139.2, 134.6, 128.0
127.8, 124.5, 117.8, 115.7, 114.4, 107.5, 107.0, 104.2, 102.0, 63.7,
4.1.2.4. Dimethyl 7-[3-(4-pyridinyl)propyl]indolizine-1,3-dicarboxylate 55.9, 55.4, 51.4, 51.2, 35.1, 34.7, 29.8; HRMS (ESI) m/z: calcd. for
(12). This compound was prepared in two steps from 1,3-bis(4-pyr C29H31N2O6 [M]+ 503.2177, obsd 503.2179.
idyl)propane
4.1.3. Bis-pyridinium salts
(1) N-Alkylation
4.1.3.1. 1-Methyl-4-[2-[1-(2-methoxy-2-oxoethyl)pyridinium-4-yl]ethyl]
1,3-Bis(4-pyridyl)propane (502 mg, 2.53 mmol) was dissolved in pyridinium bromide and iodide (18). 1-Methyl-4-[2-(4-pyridinyl)ethyl]-
CH3CN (5 mL). Methyl bromoacetate (0.9 eq.) was added and the so pyridinium iodide [64] (315 mg, 0.97 mmol) was dissolved in CH3CN (7
lution was stirred at 70–80 ◦ C for 3 h. After concentration under vac mL). Methyl bromoacetate (3 eq.) was added and the solution was
uum, water was added then the aqueous solution was washed 4 times stirred at 70–80 ◦ C for 20 h. After concentration under vacuum, the
with CH2Cl2 to remove the residual 1,3-bis(4-pyridyl)propane. After residue was washed 3 times with CH2Cl2. After drying, the bi-pyridinium
aqueous phase concentration under vacuum, the residue was washed 18 was thus isolated as a brown solid in 20% yield (246 mg).
1
with cold CH3CN. The organic solution was concentrated to give brown H NMR (400 MHz, CD3OD) δ 8.89 (d, 2H, J = 6.8 Hz), 8.84 (d, 2H, J
oil composed of a 10/1 mixture of 1-(2-methoxy-2-oxoethyl)-4-[3-(4- = 6.8 Hz), 8.17 (d, 2H, J = 6.8 Hz), 8.08 (d, 2H, J = 6.8 Hz), 5.59 (s, 2H),
pyridinyl)propyl]pyridinium bromide and 4,4′ -(1,3-propanediyl)bis[1- 4.40 (s, 3H), 3.89 (s, 3H), 3.49 (m, 4H); 13C NMR (100 MHz, D20) δ
(2-methoxy-2-oxoethyl) pyridinium] dibromide in 55% yield. The 168.0, 162.3, 160.0, 145.4, 144.8, 128.1, 127.9, 125.8, 60.3, 54.0, 47.6,
mixture was directly used in the next step. 34.4, 34.1; HRMS (ESI) m/z: calcd. for C16H19N2O2 [M− H]+ 271.1441,
obsd 271.1440 and calcd. for C16H20N2O2 [M]2+ 136.0767, obsd
(2) Indolizine formation 136.0759.
The pyridinium mixture (217 mg, 0.62 mmol) and methyl propiolate 4.1.3.2. 1-[2-(4-Methoxyphenyl)-2-oxoethyl]-4-[3-(1-methylpyridinium-
(2.1 eq.) were dissolved in anh. CH2Cl2 (5 mL). NEt3 (1.2 eq.) diluted in 4-yl)propyl]pyridinium iodide and bromide (19). 1-Methyl-4-[3-(4-pyr
2 mL of anh. CH2Cl2 was then added dropwise in the reacting mixture. idinyl)propyl]-pyridinium iodide [65] (238 mg, 0.7 mmol) was dis
After 4 h stirring at rt water was added to the solution. The organic phase solved in CH3CN (8 mL). 2-Bromo-p-methoxyacetophenone (3 eq.) was
was separated and washed 2 more times with water and brine before added and the solution was stirred at 70–80 ◦ C for 20 h. After concen
being dried on MgSO4 and concentrated under reduced pressure. The tration under vacuum, the residue was washed 3 times with CH2Cl2.
residue was then purified by flash chromatography on silica gel After drying, the bi-pyridinium 19 was thus isolated as a brown solid in
(CH2Cl2/MeOH: 98/2) to obtain indolizine 12 as an amorphous orange 93% yield.
solid in 55%. 1
H NMR (400 MHz, CD3OD) δ8.81 (d, 2H, J = 6.4 Hz), 8.78 (d, 2H, J
1
H NMR (400 MHz, CDCl3) δ 9.43 (d, 1H, J = 7.2 Hz), 8.52 (d, 2H, J = 6.8 Hz), 8.13 (d, 2H, J = 6.4 Hz), 8.11 (d, 2H, J = 9.2 Hz), 8.04 (d, 2H,
= 4.4 Hz), 8.16 (s, 1H), 7.96 (s, 1H), 7.15 (d, 2H, J = 4.4 Hz), 6.85 (d, J = 6.4 Hz), 7.15 (d, 2H, J = 8.8 Hz), 4.40 (s, 3H), 3.95 (s, 3H), 3.15 (m,
1H, J = 7.2 Hz), 3.93 (s, 6H), 2.77 (t, 2H, J = 7.6 Hz), 2.71 (t, 2H, J = 4H), 2.30 (m, 2H); 13C NMR (100 MHz, D2O) δ 190.8, 163.8, 164.4,
7.6 Hz), 2.07 (m, 2H); 13C NMR (100 MHz, CDCl3) δ 164.7, 161.5, 150.6, 162.1, 145.3, 144.6, 131.3, 128.0, 126.2, 114.8, 56.1, 47.7, 34.8, 34.5,
149.7, 140.5, 139.5, 127.7 124.5, 124.1, 123.9, 117.7, 115.9, 114.1, 28.5; HRMS (ESI) m/z: calcd. for C23H25N2O2 [M− H]+ 361.1911, obsd
104.0, 51.9, 51.3, 34.9, 34.6, 30.7; HRMS (ESI) m/z: calcd. for 361.1912 and calcd. for C23H26N2O2 [M]2+ 181.0992, obsd 181.0996.
C12H21N2O4 [M + H]+ 353.1496, obsd 353.1495.
4.1.3.3. 1-Methyl-4-[3-[1-(2-methoxy-2-oxoethyl)pyridinium-4-yl]propyl]
4.1.2.5. Dimethyl 7-[3-(1-methylpyridinium-4-yl)propyl]indolizine-1,3- pyridinium bromide and iodide (20). 1-Methyl-4-[2-(4-pyridinyl)pro
dicarboxylate iodide (15). The indolizine 12 (50 mg, 0.14 mmol) was pyl]-pyridinium iodide [65] (216 mg, 0.63 mmol) was dissolved in
12
CH3CN (5 mL). Methyl bromoacetate (2 eq.) was added and the solution parameters) curve. Results were expressed as the mean ± SD of at least
was stirred at 70–80 ◦ C for 5 h. After concentration under vacuum, the three different experiments performed in triplicate. The blanks con
residue was washed 3 times with CH2Cl2. After drying, the bi-pyridinium tained no enzyme solution, its volume being replaced by ultrapure water
20 was thus isolated as a brown solid in 80% yield (246 mg). [14].
1
H NMR (400 MHz, CD3OD) δ 8.89 (d, 2H, J = 6.8 Hz), 8.83 (d, 2H, J
= 6.8 Hz), 8.15 (d, 2H, J = 6.8 Hz), 8.06 (d, 2H, J = 6.8 Hz), 5.61 (s, 2H), 4.2.3. Molecular docking calculations
4.42 (s, 3H), 3.91 (s, 3H), 3.15 (m, 4H), 2.29 (m, 2H); 13C NMR (100 All modelling calculations were performed with the Schrödinger
MHz, D2O) δ 168.2, 164.4, 162.1, 145.2, 144.6, 128.8, 128.0, 125.2, Suite version 2017–4.
60.4, 54.2, 47.7, 34.8, 34.5, 28.5; HRMS (ESI) m/z: calcd. for
C17H21N2O2 [M− H]+ 285.1598, obsd 285.1597 and calcd. for 4.2.3.1. Protein preparation. The crystal structures of the complexes of
C17H22N2O2 [M]2+ 143.0835, obsd 143.0838. hAChE with dihydrotanshinone I (PDB ID: 4M0E [66], solved at 2 Å
resolution) and of hBChE with (S)-2-(butylamino)-N-(2-cyclo
4.2. Biological evaluation heptylethyl)-3-(1H-indol-3-yl) propanamide (PDB ID:6QAA, solved at
1.9 Å) [67], were used for the docking calculations. The proteins were
4.2.1. Antioxidant activity: DPPH assay prepared using the Protein Preparation Wizard tool [68]. Missing side-
In a 96-well plate, 100 µL of a 100 µg/mL DPPH methanolic solution chains were built and water molecules were retrieved. The partial
were added to 100 µL samples at concentrations ranging from 100 to atomic charges were derived from the OPLS3 force field [69]. Hydrogen
1500 µg/mL in methanol. The resulting solutions were then homoge atoms were added considering the pH7 value and using the algorithm
nized and left in the dark at rt for 35 min. The absorbance values at 517 PropKa [70]. The orientation of the asparagine and glutamine side-
nm were recorded after this time period. The blank sample used was a chains were performed so as to maximize the hydrogen bond network.
1:1 mixture of methanol and DPPH solution. A 48 µg/mL solution of An all-atom energy minimization with a 0.3 Å heavy-atom RMSD
rutin was used as a reference standard. The DPPH inhibition percentage, criteria for termination was performed using the Impref utility [71] and
which relates to the antioxidant activity of rutin or of the tested com OPLS3.
pounds, was calculated using the following formula: % inhibition =
(Ablank – Asample)/Ablank * 100. The assays were performed in triplicates. 4.2.3.2. Ligand preparation. The low-energy conformers of the com
The initial solutions of samples were prepared by dilution of 20 mM pounds were generated using the LigPrep module [72]. The ligands were
stock DMSO solutions with MeOH to reach the concentration of 3 mg/ docked flexibly; the parameters for van der Waals radii were scaled by
mL. For all tested compounds, the highest concentration in well was 0.80 for atoms with partial charges less than 0.15e, and non-planar
0.75 mg/mL. amide bonds were penalized.
Relevant ionization and tautomeric states at pH comprised between
4.2.2. Cholinesterase inhibition assays 5 and 9 were generated using the Hammett and Taft methodology
Acetylcholinesterase ((E.C. 3.1.1.7) from Electrophorus electricus implemented in Epik [73]. The resulting conformations were energy-
(eeAChE), Type V-S, (Sigma C2888) ≥ 1000U/mg protein; Butyr minimized using OPLS3 force field [69] and a maximum of 100 con
ylcholinesterase (E.C. 3.1.1.8) from equine serum (eqBChE) (Sigma ref formations per ligand were generated and subjected to post-docking
C1057), ≥900U/mg protein; Human acetylcholinesterase, recombinant, minimization.
expressed in HEK 293 cells, lyophilized powder, (ref Sigma C1682),
≥1000U/mg protein; Human butyrylcholinesterase (hBChE) from 4.2.3.3. Docking calculations. The Glide algorithm version 7.7 [74,75]
human serum (ref Sigma B4186), 4U/vial; 5,5′ -dithiobis-(2- nitro in Standard precision mode with the OPLS3 force field was used. The
benzoic acid) (Ellman’s reagent, DTNB), S-acetylthiocholine iodide receptors were considered as rigid, but the hydroxyls of Ser, Thr and Tyr
(ATCI), S-butyrylthiocholine iodide, and donepezil hydrochloride were residues were allowed to rotate, including residue Y337, known to adopt
purchased from Sigma Aldrich. different conformers depending on the ligand that is co-crystallized
The experimental procedure described below for AChE was also [76]. A large receptor grid encompassing both the catalytic site and
applied to BChE by replacing the S-acetylthiocholine by the S- the peripheral site was generated for both enzymes. We used a scaling of
butyrylthiocholine. van der Waals radii of 1 for atoms with partial charges less than 0.25e.
AChE solution was prepared in Tris⋅HCl buffer (pH = 8; 50 mM, with The shape and physico-chemical properties of the binding sites were
BSA (Bovine serum albumine) 0.1% w), and then diluted at an activity of mapped onto grids defined with large dimensions to reflect the presence
0.27U for non-symmetrical hybrids (the dilution was 0.5U for the of the two main binding sites. The dimensions of the receptor grids were
symmetrical bis-pyridinium salts and bis-indolizines 1′-4′ series, see respectively 36*42*35 Å3 and 31*35*30 Å3 for respectively the hAChE
Table S4). S-Acetylthiocholine iodide solution was prepared in ultrapure and hBChE enzymes, centered on the co-crystallized ligand. Our docking
water, at the concentration of 15 mM. DTNB (5,5′ -dithiobis(2-nitro procedure was validated as it was able to recover the binding pose of
benzoic acid)) was diluted in phosphate buffer (pH = 7.4; 1.5 mM). dihydrotanshinone I in hAChE with a RMSD of 0.017 Å.
The test solutions were prepared at a concentration of 20 mM (stock
solutions in DMSO) and then diluted with Tris⋅HCl buffer (pH = 8.0, 50 4.2.4. Study on amyloid fibers
mM, 0.1 M NaCl, 0.02 M MgCl2), freshly prepared solutions were used The AcPHF6 (Ac-VQIVYK-NH2) peptide was prepared as previously
for assays performed in the same day. described [46]. The AcR3 peptide corresponds to G271-G323ΔR2 tau
In 96-well plates, 160 µL of 1.5 mM DTNB solution and 50 µL of AChE fragment and was synthesized by Fmoc/tBu strategy on Rink amide
(0.5 or 0.27U/mL) solutions were incubated with 10 μL of various MBHA resin.
concentrations of the tested compounds (final concentrations in wells
were 100–0.04 µM (compared to 20–0.05 µM for the symmetrical 4.2.4.1. Thioflavin T fluorescence assays. A typical ThT assay’s micro
compounds, Table S4) at 37 ◦ C, and then 30 μL of acetylthiocholine plate includes ThT control wells, peptide (AcPHF6 or R3) control wells
iodide (15 mM) were added. The absorbance was measured at 48 s in and Ind-PyCx control wells. The wells were prepared in hexaplicate.
tervals for minimum 20 cycles at 405 nm wavelength with the Omega The total volume of each well was 100 µL for a final concentration of
fluorescence plate reader (BMG LABTECH GmbH, Offenburg, Germany). 100 µM peptide, 10 µM ThT, 10 µM heparin (for AcR3) and 100 µM Ind-
IC50 values were obtained by plotting the initial velocities versus the PyCx with a maximum of 0.5% DMSO in 50 mM phosphate buffer pH
concentrations of the tested compounds and using nonlinear regression 7.4. In blank wells, the peptide solution above was replaced by pure
analysis of the dose–response - inhibition ([inhibitor] vs response, 3
13
water. The assays were conducted in standard 96-well plates (U-shape, facilities for mass spectrometry analyses (A. Durand, L. Fort, R. Gueret)
black, polypropylene, Greiner Bio-One, USA) in a POLARstar Omega and for NMR analyses (M.-C. Molina).
fluorescence plate reader (BMG LABTECH GmbH, Offenburg, Germany)
at room temperature (AcPHF6 peptide) or 37 ◦ C (AcR3 peptide) with Appendix A. Supplementary material
excitation and emission wavelengths at 440 nm and 480 nm, respec
tively. Fluorescence data were recorded every minute over 2 h (AcPHF6 Supplementary data to this article can be found online at https://doi.
peptide) or over 34 h (AcR3 peptides) with 10 s shaking (700 rpm) prior org/10.1016/j.bioorg.2021.105390.
to each reading (top optic, orbital averaging of 3 mm diameter) with a
gain of 1200. Inhibitions reported in Fig. 6 are calculated from the
References
average of the hexaplicate, and the errors represent the SEM of three
independent experiments performed in hexaplicates. [1] S. Small, R. Mayeux, Alzheimer’s disease and related dementias, Merrit’s,
Neurology 107 (2005) 771–776.
[2] J. Cummings, G. Lee, A. Ritter, M. Sabbagh, K. Zhong, Alzheimer’s disease drug
4.2.4.2. Absorption spectra. Absorption spectra were acquired on a
development pipeline: 2019, Alzheimer’s Dement. 5 (2019) 272–293.
spectrophotometer UV-1650 PC Shimadzu using an optical length of 10 [3] P. Anand, B. Singh, A review on cholinesterase inhibitors for Alzheimer’s disease,
mm. Acquisitions were recorded from 200 nm to 600 nm. Samples were Arch. Pharm. Res. 36 (4) (2013) 375–399.
[4] M. de Freitas Silva, K.S. Dias, V.S. Gontijo, C.J.C. Ortiz, J. Viegas, Multi-target
prepared to give a final concentration of 100 µM Ind-PyCx with a
directed drugs as a modern approach for drug design towards Alzheimer’s disease:
maximum of 0.5% DMSO in 50 mM phosphate buffer pH 7.4. an update, Curr. Med. Chem. 25 (29) (2018) 3491–3525.
[5] R. Leon, A.G. Garcia, J. Marco-Contelles, Recent advances in the multitarget-
directed ligands approach for the treatment of Alzheimer’s disease, Med. Res. Rev.
4.2.4.3. Emission spectra. Emission spectra were acquired on a Fluo
33 (1) (2013) 139–189.
roMax4 Horiba Spectrometer a quartz cell with a 3 mm path length and a [6] P. Mishra, A. Kumar, G. Panda, Anti-cholinesterase hybrids as multi-target-directed
maximum volume of 300 µL. ligands against Alzheimer’s disease (1998–2018), Bioorg. Med. Chem. 27 (6)
(2019) 895–930.
Emission spectra of Ind-PyCx series alone were recorded from 100
[7] S. Montanari, M. Bartolini, P. Neviani, F. Belluti, S. Gobbi, L. Pruccoli, A. Tarozzi,
µM Ind-PyCx alone in 50 mM phosphate buffer pH 7.4 with a maximum F. Falchi, V. Andrisano, P. Miszta, A. Cavalli, S. Filipek, A. Bisi, A. Rampa,
of 0.5% DMSO. Their emission spectra in presence of pre-formed fibers Multitarget Strategy to Address Alzheimer’s Disease: Design, Synthesis, Biological
were recorded after adding the Ind-PyCx solution in 50 mM phosphate Evaluation, and Computational Studies of Coumarin-Based Derivatives,
ChemMedChem 11 (12) (2016) 1296–1308.
buffer pH 7.4 (0.5% DMSO) to preformed AcPHF6 fibers (or AcR3 fibers) [8] C.J.C. Ortiz, M. de Freitas Silva, V.S. Gontijo, F.P.D. Viegas, K.S.T. Dias, C. Viegas,
in presence or without ThT. The final concentration of Ind-PyCx com Design of Multi-target Directed Ligands as a Modern Approach for the Development
pounds in cell was 100 µM. AcPHF6 pre-formed fibers were prepared in of Innovative Drug Candidates for Alzheimer’s Disease, in: Methods in
Pharmacology and Toxicology, Springer Science, New-York, 2018, pp. 255–351.
an Eppendorf® Protein LoBind tube by incubating 100 µM AcPHF6 [9] T. Wang, X.-H. Liu, J. Guan, S. Ge, M.-B. Wu, J.-P. Lin, L.-R. Yang, Advancement of
peptide in phosphate buffer for 24 h at room temperature in absence or multi-target drug discoveries and promising applications in the field of Alzheimer’s
in presence of 10 µM ThT (for the displacement assay). For AcR3 pre- disease, Eur. J. Med. Chem. 169 (2019) 200–223.
[10] P. Zhang, S. Xu, Z. Zhu, J. Xu, Multi-target design strategies for the improved
formed fibers, they were prepared in an Eppendorf® Protein LoBind treatment of Alzheimer’s disease, Eur. J. Med. Chem. 176 (2019) 228–247.
tube by incubating 100 µM AcR3 peptide and 10 µM heparin in phos [11] M. Singh, M. Kaur, N. Chadha, O. Silakari, Hybrids: a new paradigm to treat
phate buffer for 24 h at 37 ◦ C in absence or presence of 10 µM ThT (for Alzheimer’s disease, Mol. Divers. 20 (1) (2016) 271–297.
[12] M. Alipour, M. Khoobi, A. Foroumadi, H. Nadri, A. Moradi, A. Sakhteman,
the displacement assay).
M. Ghandi, A. Shafiee, Novel coumarin derivatives bearing N-benzyl pyridinium
moiety: potent and dual binding site acetylcholinesterase inhibitors, Bioorg. Med.
4.2.4.4. Circular dichroism. CD spectra were acquired on a JASCO J-810 Chem. 20 (24) (2012) 7214–7222.
[13] F. Baharloo, M.H. Moslemin, H. Nadri, A. Asadipour, M. Mahdavi, S. Emami,
spectropolarimeter (Jasco Corporation, Tokyo, Japan) using a quartz L. Firoozpour, R. Mohebat, A. Shafiee, A. Foroumadi, Benzofuran-derived
cell with a 1 mm path length and a maximum volume of 300 µL. Ac benzylpyridinium bromides as potent acetylcholinesterase inhibitors, Eur. J. Med.
quisitions were recorded at room temperature between 260 nm and 206 Chem. 93 (2015) 196–201.
[14] J.-S. Lan, T. Zhang, Y. Liu, J. Yang, S.-S. Xie, J. Liu, Z.-Y. Miao, Y. Ding, Design,
nm with a spectral resolution of 1 nm, an average time of 1 s per data
synthesis and biological activity of novel donepezil derivatives bearing N-benzyl
point and a band width of 10 nm. The spectra reported represent an pyridinium moiety as potent and dual binding site acetylcholinesterase inhibitors,
average of four acquisitions and have not be smoothed. The total volume Eur. J. Med. Chem. 133 (2017) 184–196.
[15] R. Palin, J.K. Clark, P. Cowley, A.W. Muir, E. Pow, A.B. Prosser, R. Taylor, M.-
of each quartz cell was 300 µL for a final concentration of 100 µM AcR3,
Q. Zhang, Novel piperidinium and pyridinium agents as water-soluble
10 µM heparin and 100 µM Ind-PyCx in 50 mM phosphate buffer pH 7.4 acetylcholinesterase inhibitors for the reversal of neuromuscular blockade, Bioorg.
with a maximum of 0.5% DMSO. Prior to the acquisitions in the quartz Med. Chem. Lett. 12 (18) (2002) 2569–2572.
cell, the mixtures were incubated at 37 ◦ C for 24 h in an Eppendorf® [16] F. Vafadarnejad, E. Karimpour-Razkenari, B. Sameem, M. Saeedi, O. Firuzi,
N. Edraki, M. Mahdavi, T. Akbarzadeh, Novel N-benzylpyridinium moiety linked to
Protein LoBind tube. arylisoxazole derivatives as selective butyrylcholinesterase inhibitors: Synthesis,
biological evaluation, and docking study, Bioorg. Chem. 92 (2019), 103192.
[17] M. Khoobi, M. Alipour, A. Sakhteman, H. Nadri, A. Moradi, M. Ghandi, S. Emami,
Declaration of Competing Interest A. Foroumadi, A. Shafiee, Design, synthesis, biological evaluation and docking
study of 5-oxo-4, 5-dihydropyrano [3, 2-c] chromene derivatives as
acetylcholinesterase and butyrylcholinesterase inhibitors, Eur. J. Med. Chem. 68
The authors declare that they have no known competing financial (2013) 260–269.
interests or personal relationships that could have appeared to influence [18] H. Akrami, B.F. Mirjalili, M. Khoobi, H. Nadri, A. Moradi, A. Sakhteman, S. Emami,
the work reported in this paper. A. Foroumadi, A. Shafiee, Indolinone-based acetylcholinesterase inhibitors:
synthesis, biological activity and molecular modeling, Eur. J. Med. Chem. 84
(2014) 375–381.
Acknowledgments [19] M. Mollazadeh, M. Mohammadi-Khanaposhtani, A. Zonouzi, H. Nadri, Z. Najafi,
B. Larijani, M. Mahdavi, New benzyl pyridinium derivatives bearing 2, 4-dioxo
chroman moiety as potent agents for treatment of Alzheimer’s disease: Design,
This work has been partially supported by Labex Arcane, CBH-EUR- synthesis, biological evaluation, and docking study, Bioorg. Chem. 87 (2019)
GS (ANR-17-EURE-0003), by the PolyNatCarnot Institute (Investisse 506–515.
[20] S. Sowmiah, J.M. Esperança, L.P. Rebelo, C.A. Afonso, Pyridinium salts: from
ments d’Avenir - grant agreement n◦ ANR-11-CARN-007-01), by Neu
synthesis to reactivity and applications, Org. Chem. Front. 5 (3) (2018) 453–493.
roCoG IDEX UGA in the framework of the “Investissements d’avenir” [21] P. Sharma, P. Srivastava, A. Seth, P.N. Tripathi, A.G. Banerjee, S.K. Shrivastava,
program (ANR-15-IDEX-02), and by a grant of the Romanian National Comprehensive review of mechanisms of pathogenesis involved in Alzheimer’s
Authority for Scientific Reasearch and Innovation (CNCS-UEFISCDI, PN- disease and potential therapeutic strategies, Prog. Neurobiol. 174 (2019) 53–89.
[22] I.O. Ghinea, R.M. Dinica, Breakthroughs in Indole and Indolizine Chemistry–New
II-ID-PCE-2011-3-0226, CCCDI-UEFISCDI, project number 85BM/ Synthetic Pathways, New Applications, Scope of Selective Heterocycles from
2017). The NanoBio ICMG (UAR 2607) is acknowledged for providing Organic and Pharmaceutical Perspective, IntechOpen, eBook, 2016.
14
[23] V. Sharma, V. Kumar, Indolizine: a biologically active moiety, Med. Chem. Res. 23 [48] J. Zheng, C. Liu, M.R. Sawaya, B. Vadla, S. Khan, R.J. Woods, D. Eisenberg, W.
(8) (2014) 3593–3606. J. Goux, J.S. Nowick, Macrocyclic β-sheet peptides that inhibit the aggregation of a
[24] M. Arvin-Berod, A. Desroches-Castan, S. Bonte, S. Brugière, Y. Coute, L. Guyon, J.- tau-protein-derived hexapeptide, J. Am. Chem. Soc. 133 (9) (2011) 3144–3157.
J. Feige, I. Baussanne, M. Demeunynck, Indolizine-based scaffolds as efficient and [49] W.J. Goux, L. Kopplin, A.D. Nguyen, K. Leak, M. Rutkofsky, V.
versatile tools: Application to the synthesis of biotin-tagged antiangiogenic drugs, D. Shanmuganandam, D. Sharma, H. Inouye, D.A. Kirschner, The formation of
ACS omega 2 (12) (2017) 9221–9230. straight and twisted filaments from short tau peptides, J. Biol. Chem. 279 (26)
[25] S.-K. Choi, J. Rho, S.E. Yoon, J.-H. Seok, H. Kim, J. Min, W. Yoon, S. Lee, H. Yun, (2004) 26868–26875.
O.-P. Kwon, Full Color Tunable Aggregation-Induced Emission Luminogen for [50] H. Kalász, S. Nurulain, G. Veress, S. Antus, F. Darvas, E. Adeghate, A. Adem,
Bioimaging Based on an Indolizine Molecular Framework, Bioconjug. Chem. 31 F. Hashemi, K. Tekes, Mini review on blood–brain barrier penetration of
(11) (2020) 2522–2532. pyridinium aldoximes, J. Appl. Toxicol. 35 (2) (2015) 116–123.
[26] D. Kim, J.H. Lee, H.Y. Kim, J. Shin, K. Kim, S. Lee, J. Park, J. Kim, Y. Kim, [51] A. Tsuji, Small molecular drug transfer across the blood-brain barrier via carrier-
Fluorescent indolizine derivative YI-13 detects amyloid-β monomers, dimers, and mediated transport systems, NeuroRx 2 (1) (2005) 54–62.
plaques in the brain of 5XFAD Alzheimer transgenic mouse model, PloS one 15 [52] T.T. Wager, R.Y. Chandrasekaran, X. Hou, M.D. Troutman, P.R. Verhoest,
(12) (2020), e0243041. A. Villalobos, Y. Will, Defining desirable central nervous system drug space
[27] J. Sung, Y. Lee, J.-H. Cha, S.B. Park, E. Kim, Development of fluorescent through the alignment of molecular properties, in vitro ADME, and safety
mitochondria probe based on 1, 2-dihydropyrrolo [3, 4-b] indolizine-3-one, Dyes attributes, ACS Chem. Neurosci. 1 (6) (2010) 420–434.
Pigments 145 (2017) 461–468. [53] H. Pajouhesh, G.R. Lenz, Medicinal chemical properties of successful central
[28] N.-L. Marangoci, L. Popovici, E.-L. Ursu, R. Danac, L. Clima, C. Cojocaru, nervous system drugs, NeuroRx 2 (4) (2005) 541–553.
A. Coroaba, A. Neamtu, I. Mangalagiu, M. Pinteala, A. Rotaru, Pyridyl-indolizine [54] The bis-pyridinium di-cations 17-20 (and their symmetrical analogs 1’a-c and 2’a-
derivatives as DNA binders and pH-sensitive fluorescent dyes, Tetrahedron 72 (50) c, see Table S4) are too polar (strongly negative LogP). It is also interesting to
(2016) 8215–8222. notice that the bis-indolizines (7,7’-bis-indolizines 3’a-c and bis(indolizin-7-yl)
[29] X. Zheng, R. Ji, X. Cao, Y. Ge, FRET-based ratiometric fluorescent probe for Cu2+ ethane 4’a-c) do not in general follow the rules as they are too lipophilic (LogP >
with a new indolizine fluorophore, Anal. Chim. Acta 978 (2017) 48–54. 6), showing both high MW and PSA values (see Table S4). Their calculated high
[30] W.E. Meador, S.A. Autry, R.N. Bessetti, J.N. Gayton, A.S. Flynt, N.I. Hammer, J. lipophilicity is experimentally correlated to a low solubility in water with
H. Delcamp, Water-soluble NIR absorbing and emitting indolizine cyanine and formation of aggregates.
indolizine squaraine dyes for biological imaging, J. Org. Chem. 85 (6) (2020) [55] G.L. Ellman, K.D. Courtney, V. Andres, R.M. Featherstone, A new and rapid
4089–4095. colorimetric determination of acetylcholinesterase activity, Biochem. Pharmacol. 7
[31] H. Chakravarty, Y. Ju, W.H. Chen, K.Y. Tam, Dual targeting of cholinesterase and (2) (1961) 88–95.
amyloid beta with pyridinium/isoquinolium derivatives, Drug Dev. Res. 81 (2) [56] Q. Shen, B. Zhang, R. Xu, Y. Wang, X. Ding, P. Li, Antioxidant activity in vitro of
(2020) 242–255. the selenium-contained protein from the Se-enriched Bifidobacterium animalis 01,
[32] M. Abdullaha, V.K. Nuthakki, S.B. Bharate, Discovery of methoxy-naphthyl linked Anaerobe 16 (4) (2010) 380–386.
N-(1-benzylpiperidine) benzamide as a blood-brain permeable dual inhibitor of [57] H. Levine III, Thioflavine T interaction with synthetic Alzheimer’s disease
acetylcholinesterase and butyrylcholinesterase, Eur. J. Med. Chem. 207 (2020), β-amyloid peptides: Detection of amyloid aggregation in solution, Protein Sci. 2 (3)
112761. (1993) 404–410.
[33] V. Hrabcova, J. Korabecny, B. Manyova, L. Matouskova, T. Kucera, R. Dolezal, [58] K. Gade Malmos, L.M. Blancas-Mejia, B. Weber, J. Buchner, M. Ramirez-Alvarado,
K. Musilek, L. Gorecki, E. Nepovimova, K. Kuca, Bis-isoquinolinium and bis- H. Naiki, D. Otzen, ThT 101: a primer on the use of thioflavin T to investigate
pyridinium acetylcholinesterase inhibitors: in vitro screening of probes for novel amyloid formation, Amyloid 24 (1) (2017) 1–16.
selective insecticides, RSC Adv. 7 (62) (2017) 39279–39291. [59] L.P. Jameson, N.W. Smith, S.V. Dzyuba, Dye-binding assays for evaluation of the
[34] R. Azzouz, L. Peauger, V. Gembus, M.-L. Ţînţaş, J. Sopková-de Oliveira Santos,, effects of small molecule inhibitors on amyloid (Aβ) self-assembly, ACS Chem.
C. Papamicaël, V. Levacher, Novel donepezil-like N-benzylpyridinium salt Neurosci. 3 (11) (2012) 807–819.
derivatives as AChE inhibitors and their corresponding dihydropyridine “bio- [60] S.A. Hudson, H. Ecroyd, T.W. Kee, J.A. Carver, The thioflavin T fluorescence assay
oxidizable” prodrugs: Synthesis, biological evaluation and structure-activity for amyloid fibril detection can be biased by the presence of exogenous
relationship, Eur. J. Med. Chem. 145 (2018) 165–190. compounds, FEBS J. 276 (20) (2009) 5960–5972.
[35] F. Hosseini, A. Ramazani, M. Mohammadi-Khanaposhtani, M.B. Tehrani, H. Nadri, [61] J. Yang, J. Guo, J. Yuan, In vitro antioxidant properties of rutin, Food Sci. Technol.
B. Larijani, M. Mahdavi, Design, synthesis, and biological evaluation of novel 4- 41 (6) (2008) 1060–1066.
oxobenzo [d] 1, 2, 3-triazin-benzylpyridinum derivatives as potent anti-Alzheimer [62] For comparison the evaluation of the symmetrical analogs was made (see Table S4)
agents, Bioorg. Med. Chem. 27 (13) (2019) 2914–2922. using eeAChE at a concentration of 0.5U. Compounds 1’a-b, 2’a, 2’c, 3’a-b
[36] M. Singh, M. Kaur, H. Kukreja, R. Chugh, O. Silakari, D. Singh, exhibited micromolar IC50 values (2-5 µM).
Acetylcholinesterase inhibitors as Alzheimer therapy: from nerve toxins to [63] B.M. Liederer, R.T. Borchardt, Enzymes involved in the bioconversion of ester-
neuroprotection, Eur. J. Med. Chem. 70 (2013) 165–188. based prodrugs, J. Pharm. Sci. 95 (6) (2006) 1177–1195.
[37] O.B. Østby, B. Dalhus, L.L. Gundersen, F. Rise, A. Bast, G.R. Haenen, Synthesis of 1- [64] B. Furdui, O. Constantin, A. Tabacaru, R. Dinica, New bis-pyridinium diquaternary
Substituted 7-Cyano-2, 3-diphenylindolizines and Evaluation of Antioxidant salts with antimicrobial properties, Rev. Chim 63 (2012) 667–671.
Properties, Eur. J. Org. Chem. 2000 (22) (2000) 3763–3770. [65] D.H. Taffa, M. Kathiresan, L. Walder, B. Seelandt, M. Wark, Pore size and surface
[38] L.-L. Gundersen, K.E. Malterud, A.H. Negussie, F. Rise, S. Teklu, O.B. Østby, charge control in mesoporous TiO 2 using post-grafted SAMs, Phys. Chem. Chem.
Indolizines as novel potent inhibitors of 15-lipoxygenase, Bioorg. Med. Chem. 11 Phys. 12 (7) (2010) 1473–1482.
(24) (2003) 5409–5415. [66] J. Cheung, E.N. Gary, K. Shiomi, T.L. Rosenberry, Structures of human
[39] G. Ghotbi, M. Mahdavi, Z. Najafi, F.H. Moghadam, M. Hamzeh-Mivehroud, acetylcholinesterase bound to dihydrotanshinone I and territrem B show peripheral
S. Davaran, S. Dastmalchi, Design, synthesis, biological evaluation, and docking site flexibility, ACS Med. Chem. Lett. 4 (11) (2013) 1091–1096.
study of novel dual-acting thiazole-pyridiniums inhibiting acetylcholinesterase and [67] A. Meden, D. Knez, M. Jukič, X. Brazzolotto, M. Gršič, A. Pišlar, A. Zahirović,
β-amyloid aggregation for Alzheimer’s disease, Bioorg. Chem. 103 (2020), 104186. J. Kos, F. Nachon, J. Svete, Tryptophan-derived butyrylcholinesterase inhibitors as
[40] S. Abdpour, L. Jalili-Baleh, H. Nadri, H. Forootanfar, S.N.A. Bukhari, A. Ramazani, promising leads against Alzheimer’s disease, Chem. Commun. 55 (26) (2019)
S.E.S. Ebrahimi, A. Foroumadi, M. Khoobi, Chromone derivatives bearing 3765–3768.
pyridinium moiety as multi-target-directed ligands against Alzheimer’s disease, [68] G.M. Sastry, M. Adzhigirey, T. Day, R. Annabhimoju, W. Sherman, Protein and
Bioorg. Chem. 110 (2021), 104750. ligand preparation: parameters, protocols, and influence on virtual screening
[41] M. Ahmed, J.B.T. Rocha, C.M. Mazzanti, A.L. Morsch, D. Cargnelutti, M. Corrêa, enrichments, J. Comput. Aided Mol. Des. 27 (3) (2013) 221–234.
V. Loro, V.M. Morsch, M.R. Schetinger, Malathion, carbofuran and paraquat inhibit [69] E. Harder, W. Damm, J. Maple, C. Wu, M. Reboul, J.Y. Xiang, L. Wang, D. Lupyan,
Bungarus sindanus (krait) venom acetylcholinesterase and human serum M.K. Dahlgren, J.L. Knight, OPLS3: a force field providing broad coverage of drug-
butyrylcholinesterase in vitro, Ecotoxicol. 16 (4) (2007) 363. like small molecules and proteins, J. Chem. Theory Comput. 12 (1) (2016)
[42] P. Kapková, V. Alptüzün, P. Frey, E. Erciyas, U. Holzgrabe, Search for dual function 281–296.
inhibitors for Alzheimer’s disease: Synthesis and biological activity of [70] H. Li, A.D. Robertson, J.H. Jensen, Very fast empirical prediction and
acetylcholinesterase inhibitors of pyridinium-type and their Aβ fibril formation rationalization of protein pKa values, Proteins 61 (4) (2005) 704–721.
inhibition capacity, Bioorg. Med. Chem. 14 (2) (2006) 472–478. [71] J.L. Banks, H.S. Beard, Y. Cao, A.E. Cho, W. Damm, R. Farid, A.K. Felts, T.
[43] Y. Seto, T. Shinohara, Structure-activity relationship of reversible cholinesterase A. Halgren, D.T. Mainz, J.R. Maple, Integrated modeling program, applied
inhibitors including paraquat, Arch. Toxicol. 62 (1) (1988) 37–40. chemical theory (IMPACT), J. Comput. Chem. 26 (16) (2005) 1752–1780.
[44] B. Furdui, R. Dinica, M. Demeunynck, I. Druta, A. Vlahovici, New reactive [72] LigPrep, LigPrep version 3.5 Schrödinger, LLC, New York, NY, 2015. 2015.
pyridinium-indolizines fluorophores, Rev. Roum. Chim. 52 (7) (2007) 633–637. [73] Epik, Epik version 4.2 Schrödinger, LLC, New York, NY, 2017. 2017.
[45] B. Furdui, R. Dinica, I.I. Druta, M. Demeunynck, Improved synthesis of cationic [74] R.A. Friesner, J.L. Banks, R.B. Murphy, T.A. Halgren, J.J. Klicic, D.T. Mainz, M.
pyridinium-substituted indolizines, Synthesis 2006 (16) (2006) 2640–2642. P. Repasky, E.H. Knoll, M. Shelley, J.K. Perry, Glide: a new approach for rapid,
[46] L. Lunven, H. Bonnet, S. Yahiaoui, W. Yi, L.N. Da Costa, M. Peuchmaur, accurate docking and scoring. 1. Method and assessment of docking accuracy,
A. Boumendjel, S. Chierici, Disruption of fibers from the tau model AcPHF6 by J. Med. Chem. 47 (7) (2004) 1739–1749.
naturally occurring aurones and synthetic analogues, ACS Chem. Neurosci. 7 (7) [75] Glide, Glide version 7.7, Schrödinger, LLC, New York, NY, 2017. 2017.
(2016) 995–1003. [76] A. Hameed, S.T. Zehra, S.J. Shah, K.M. Khan, R.D. Alharthy, N. Furtmann,
[47] T. Mohamed, T. Hoang, M. Jelokhani-Niaraki, P.P. Rao, Tau-derived-hexapeptide J. Bajorath, M.N. Tahir, J. Iqbal, Syntheses, cholinesterases inhibition, and
306VQIVYK311 aggregation inhibitors: nitrocatechol moiety as a pharmacophore molecular docking studies of Pyrido [2, 3-b] pyrazine derivatives, Chem. Biol. Drug
in drug design, ACS Chem. Neurosci. 4 (12) (2013) 1559–1570. Des. 86 (5) (2015) 1115–1120.
15

Molecular Docking - Indolizine

Uploaded by

Copyright:

Available Formats

Molecular Docking - Indolizine

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Molecular Docking - Indolizine

Uploaded by

Copyright:

Available Formats

Bioorganic Chemistry 116 (2021) 105390

Contents lists available at ScienceDirect

Interest of novel N-alkylpyridinium-indolizine hybrids in the field of

Fig. 1. Examples of pyridinium salts and indolizine derivatives of biological interest.

Table 1 the tested enzymes, we wondered if these esters could be substrates of

You might also like