Potential prodrugs of non-steroidal anti-inflammatory agents for targeted drug delivery to the CNS

Article abstract of DOI:10.1016/j.ejmech.2004.05.006

Recently non-steroidal anti-inflammatory drugs (NSAIDs) have been proposed to prevent or to cure Alzheimer's disease. In this respect, we synthesized new potential prodrugs of several NSAIDs in order to increase their access to the brain. The carboxylic g

Full text of DOI:10.1016/j.ejmech.2004.05.006

European Journal of Medicinal Chemistry 39 (2004) 715–727
www.elsevier.com/locate/ejmech
Original article
Potential prodrugs of non-steroidal anti-inflammatory agents for targeted
drug delivery to the CNS
Luana Perioli *, Valeria Ambrogi, Claudia Bernardini, Giuliano Grandolini, Maurizio Ricci,
Stefano Giovagnoli, Carlo Rossi
Dipartimento di Chimica e Tecnologia del Farmaco, Via del Liceo 1, Università degli Studi di Perugia, 06123 Perugia, Italy
Received 27 January 2004; received in revised form 29 April 2004; accepted 10 May 2004
Abstract
Recently non-steroidal anti-inflammatory drugs (NSAIDs) have been proposed to prevent or to cureAlzheimer’s disease. In this respect, we
synthesized new potential prodrugs of several NSAIDs in order to increase their access to the brain. The carboxylic group of NSAIDs was
attached to the 1,4-dihydro-1-methylpyridine-3-carboxylate moiety, which acts as a carrier, via an amino alcohol bridge, according to the
chemical delivery approach developed by Bodor. The lipophilicity of potential prodrugs was evaluated both via traditional experimental
parameters, such as partition coefficient and chromatographic Rm value, and by predictive computational methods. From experimental
parameters, all prodrugs were more lipophilic when compared to their corresponding parent compounds and consequently a better blood brain
barrier (BBB) penetration is hypothesised. Prodrug lipophilicity was correlated with a calculated log P value according to Kowwin’s method.
The correlation between experimental Rm0 and calculated log P, determined by PLS analysis, was good for all compounds with the exception
of compound 7i. In addition the BBB permeation profile of our synthesized compounds was predicted using the BBBVolSurf model and seven
of the synthesized prodrugs resulted in good candidates for BBB penetration.
© 2004 Elsevier SAS. All rights reserved.
Keywords: Prodrugs; NSAIDs; Alzheimer’s disease; Lipophilicity; Predicting blood brain barrier permeation
1. Introduction
AD, extended clinical observations at the cellular level have
been carried out. Numerous studies indicate the involvement
of immune and chronic inflammatory mechanisms in AD
[9,10].
Alzheimer’s disease (AD) is the most common form of
dementia [1]. Although at present the aetiology of AD is not
well understood, many neurobiological and environmental
factors contributing to the pathogenesis of AD have been
described. Recently, it has been revealed that there is a
presence of an inflammatory component in the pathogenesis
of AD [1,2]. In fact, immunochemistry completed on post-
mortem AD brains revealed that numerous inflammatory
components are associated with neuritic plaques [3] and
epidemiological studies have shown that therapy with anti-
inflammatory drugs reduces the risk of developingAD [4–8].
In order to understand the mechanisms by which NSAIDs
can protect the nervous system from the ravages caused by
If the evidence for an inflammatory reaction in AD is
clear, it is still unclear whether this inflammatory reaction is
the cause of AD or only a secondary event that develops the
degenerative process [1]. All these data indicate that an
anti-inflammatory therapy [11,12] reduces the risk of devel-
oping AD in subjects without genetic predisposition, delays
the onset of AD in genetically vulnerable individuals and
delays the progressive cognitive deterioration inAD patients.
The aim of this study is to synthesize and predict the
permeation profiles of a few NSAID derivatives designed to
increase their access to the brain.
Some potential prodrugs of several NSAIDs, such as di-
clofenac (DIK), ibuprofen (IBU), ketoprofen (KET), tiapro-
fenic acid (TIAP) and tolmetin (TOLM), were synthesized
using as a carrier, the 1,4-dihydro-1-methylpyridine-3-
carboxylate, which was attached to the drug via an amino
* Corresponding author. Tel.: +39-75-585-5133; fax: +39-75-585-5163.
E-mail address: luanaper@unipg.it (L. Perioli).
© 2004 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2004.05.006

716
L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
alcohol bridge. Furthermore, in order to evaluate the effect of
the bridge on prodrug lipophilicity, different amino alcoholic
chains were introduced.
The lipophilicity of all potential prodrugs was evaluated
using both traditional experimental parameters, such as par-
tition coefficient and chromatographic Rm value, and predic-
tive computational methods. CNS penetration was predicted
using a BBB model [13] as described in the VolSurf proce-
dure [14].
Compounds 5a–q were N-methylated by a solution of
methyl iodide in nitromethane, to give the quaternary ammo-
nium salts 6a–q. These compounds were subsequently re-
duced, in dilute alkaline sodium bicarbonate solution, by
sodium dithionite, a selective reduction agent, into the 1,4-
dihydropyridine ring. This reaction was performed at room
temperature in a biphasic system of water and ethyl acetate,
in the absence of oxygen and keeping the pH of the medium
above 7.
All compounds were characterized by elemental analyses
1
and H nuclear magnetic resonance. UV spectra were re-
corded for compounds 7a–q and revealed the formation of
2. Results and discussion
1,4-dihydropyridine derivatives by their typical UV maxi-
1
mum absorption at ca. k 360 nm. The H-NMR spectra of
2.1. Chemical delivery system
7a–q showed the disappearance of typical signals due to the
3-substituted pyridine ring and revealed the presence of ap-
propriate signals due to the 1,4-dihydropyridine moiety.
Several techniques to obtain improved access of drugs to
the brain were developed. Among them, the chemical deliv-
ery system (CDS) approach developed by Bodor [15] is one
of the most interesting procedures to deliver drugs in a
sustained and specific manner to the CNS. It involves the
release of active species from a lipophilic prodrug through a
multistep conversion.
The CDS approach is based on a dihydropyridine py-
ridium salt equilibrium type redox molecular carrier, similar
to the endogenous NADH/NAD+ coenzyme system.
The CDS consists of a drug, a carrier (dihydropyridine)
and a linkage between them. A non-steroidal anti-
inflammatory drug containing a carboxylic group can be
attached to the carrier via an amino alcohol bridge. By
intravenous administration, the dihydropyridine form of the
CDS is rapidly distributed throughout the body, including
CNS. Next, the dihydropyridine moiety is oxidized to the
pyridinium salt. This hydrophilic form is sequestrated in the
CNS, whereas it is rapidly eliminated from the periphery. By
enzymatic hydrolysis, the active drug is released into the
brain and can exert its action, whilst the small carrier mol-
ecule (pyridinium salt) is actively transported out of the CNS.
This kind of drug delivery, sustained and specific, cannot be
achieved by the use of simple lipophilic drugs.
2.3. Lipophilicity
Lipophilicity is an essential feature for the penetration of a
molecule through the BBB. Thus, the lipophilicity of com-
pounds 7a–q was determined and was compared to that of the
parent drugs. First the determination of lipophilicity was
performed using the shake-flask method between an aqueous
buffer at pH 7.5 and n-octanol, followed by UV prodrug
determination after shaking. In the case of prodrugs, this
method failed because their lipophilicity increased to such an
extent that the molecule concentrations in the aqueous phase
could not be determined. Thus, the chromatographic Rm
values, related to the partition coefficient between the mobile
and stationary phase of a chromatographic system, were
determined and correlated to the lipophilicity of the com-
pounds [17,18].
The Rm values of each drug and the corresponding pro-
drugs were determined using the formula Rm = log[(1/Rf) –
1]. The Rf values were measured by means of a reverse-phase
TLC method with various concentrations of acetone and
aqueous phosphate buffer at pH 7 as a mobile phase. The Rm
values were plotted as a function of acetone concentrations
and eventually the Rm values, corresponding to a mobile
0
phase of 0% acetone R , were determined by extrapola-
2.2. Chemistry
͑
͒
m
tion of the linear part of Rm value curves to a mobile phase of
0
The synthetic route for these prodrugs 7a–q is shown in
Scheme 1. This procedure involves the formation of an ester
group between the carboxylic group of the drug and the
alcoholic function of the pyridine derivative 1–4, the succes-
sive quaternization of the pyridine nitrogen and eventually
their reduction to 1,4-dihydroderivatives.
The hydroxyalkylnicotinamides 1–4 were prepared by
reacting of the appropriate amino alcohol with the ethyl ester
of nicotinic acid at room temperature [16]. The formation of
esters 5a–q was achieved by reacting the drug with the
appropriate hydroxyalkyl nicotinamides 1–4 in acetonitrile
at room temperature in the presence of dicyclohexylcarbodi-
imide and 4-dimethylaminopyridine.
0% acetone (Table 1). Rm values were considered as lipo-
philic indexes.
All the prodrugs were found much more lipophilic when
compared to their corresponding drugs. Lipophilicity in-
creased significantly by introducing a branched substituent
(methyl or ethyl) in the ethylic bridge and by increasing the
length of the alkyl bridge (propyl > ethyl). Compounds 7g
and 7m resulted the most lipophilic compounds.
From the obtained Rm0 values, for all synthesized prodrugs
a better BBB penetration than the parent compounds may be
hypothesised. The lipophilicity of all prodrugs was also cor-
related to the calculated log P value by means of the Kow-
win’s method [19]. The correlation between experimental

L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
717
(chain)
+
HO
NH2
EtOOC
N
chain
CH CH
-
-
1
2
-
2
2
HO-(chain)-NHCO
N
-
-
CHCH2
CH3
1 - 4
D-COOH
-
-
3
4
CH2CH
CH2CH3
D-COO-(chain)-NHCO-
5 a-q
CH3I
N
-
-
CH2CH2CH2
D-COO-(chain)-NHCO-
I-
N
+
6 a-q
CH3
D-COO-(chain)-NHCO-
N
CH3
7 a-q
Cl
Cl
H2C
NH
CH3
CH
O
C
H3C
,
,
H3C
D =
CHCH2
H3C
CH2
N
CH3
IBU
TOLM
DIK
O
O
C
S
,
C
CH
CH
CH3
CH3
KET
TIAP
Scheme 1. Synthesis of prodrugs 7a–q.
0
from 3D-molecular fields to estimate the BBB permeation
for a large set of compounds.
Rm and calculated log P was determined by PLS analysis
0
[20] (Table 1). Fig. 1 shows the correlation between Rm
The model distinguished well between the BBB+ (pen-
etrating) and BBB– (non-penetrating) compounds, and was
mainly driven by a passive diffusion mechanism. The model
correctly assigned the corresponding BBB profile to more
than 90% of the compounds. Since the prediction error
(SDEP) of the discriminant PLS was 0.6 units, a confidence
interval was built in the t1–t2 space between the BBB+ and
BBB– regions. In this interval, the BBB prediction given by
the model were considered borderline and doubtful.
The BBB VolSurf model was used in this work to predict
the permeation profile of our synthesized prodrugs. All
17 compounds were built in 3D coordinates (see Section 4)
and then projected in the VolSurf model. The results of the
variations and calculated log P variations and makes clear
that a good correlation does exist, with the exception of
compound 7i (not reported in Fig. 1), which can be consid-
ered an outlier.
2.4. Predicting BBB permeation studies
The BBB is a complex cellular system and its function is
to maintain the homeostasis of the CNS by separating the
brain from the systemic blood circulation.
Crivori et al. [13] reported a model for predicting BBB
permeation of molecules based on literature data. The study
was conducted to demonstrate values of descriptors, derived

718
L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
Table 1
Log P, Rm0 and PLS data of drugs and compounds 7a–q
0
Rm (r) a
Compound
Drug
IBU
Chain
Log P (Kowwin)
3.79
PLS
1.025
1.28
–
–
1.025 (0.998)
1.281 (0.997)
0.6179 (0.990)
0.458 (0.993)
–
DIK
–
4.02
KET
TIAP
TOLM
DIK
–
3.0
–
2.82
–
–
2.02
–
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
7q
CH2CH2
4.61
2.380 (0.995)
3.663 (0.997)
–
2.38
3.66
–
IBU
CH2CH2
4.38
TOLM
KET
TIAP
DIK
CH2CH2
2.61
CH2CH2
3.59
–
–
CH2CH2
3.41
–
–
CH(CH3)CH2
CH(CH3)CH2
CH(CH3)CH2
CH(CH3)CH2
CH2CH(CH2CH3)
CH2CH(CH2CH3)
CH2CH(CH2CH3)
CH2CH(CH2CH3)
CH2CH2CH2
CH2CH2CH2
CH2CH2CH2
CH2CH2CH2
5.02
3.89 (0.991)
4.972(0.994)
3.89 (0.992)
4.57 (0.993)
4.52 (0.991)
4.269 (0.996)
3.92 (0.998)
4.96 (0.997)
3.841 (0.997)
4.082 (0.994)
4.122 (0.991)
2.91 (0.991)
4.04
4.97
–
IBU
4.80
KET
TIAP
DIK
4.01
3.83
4.0
4.52
4.27
3.93
4.26
3.84
4.08
2.88
2.58
5.51
IBU
5.29
KET
TIAP
DIK
4.50
4.32
5.10
IBU
4.87
KET
TIAP
4.08
3.90
a r = correlation coefficient.
projection are reported in Fig. 2, which shows the BBB
profile as predicted by VolSurf. All prodrugs are located in
the BBB+ region of the plot. However, from a statistical point
of view, the seven compounds coded with a, b, g, j, k, n, o
are better candidates for BBB penetration than the other
compounds located in the borderline regions of the plot.
3. Conclusions
All prodrugs were successfully synthesized and showed a
higher lipophilicity when compared to their parent drugs. In
particular lipophilicity increased significantly by introducing
a branched substituent (methyl or ethyl) in the ethyl bridge or
by increasing the length of the alkyl bridge (propyl > ethyl)
as in the case of compound 7j, a prodrug of diclofenac. All
compounds, with the exception of 7i, showed good correla-
tion between Rm0 parameters and calculated log P.
Predictive BBB permeation studies demonstrated that all
potential prodrugs are located in the BBB+ region of the plot.
In particular compounds 7a, 7b, 7g, 7j, 7k, 7n, and 7o are
the best candidates for BBB penetration. They have two
things in common, first they are prodrugs of ibuprofen and
diclofenac, and second, they all contain either a propylic or a
branched chain bridge.
4. Experimental
4.1. Chemistry
4.1.1. Materials and methods
Diclofenac sodium salt was purchased from Sigma
Chemical Company, (Milano, Italy), ibuprofen (acid), tol-
metin (acid), tiaprofenic acid and ketoprofen (acid) were
kindly supplied by Francesco Angelini S.p.A. (Roma, Italy),
Cilag (Schaffhausen, Switzerland), Camillo Corvi (Milan,
Italy) and Rhone-Poulenc Rorer (Origgio, Italy), respec-
tively. All starting materials were of reagent grade.
Fig. 1. The line represents the PLS regression model between the computed
log P and the experimental R0m values for compounds 7a–q.

L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
719
Fig. 2. BBB profile as predicted by Volsurf. a = 7a, b = 7b, c = 7c, d = 7d, e = 7e, f = 7f, g = 7g, h = 7h, i = 7i, j = 7j, k = 7k, l = 7l, m = 7m, n = 7n, o = 7o, p = 7p,
q = 7q.
1
Melting points were obtained using a Kofler-hot stage
apparatus and are uncorrected. Mass spectra were recorded
in a Varian MAT instrument. Proton nuclear magnetic reso-
nance spectra (1H-NMR) were recorded with a Bruker AC
200 (200 MHz) spectrometer, using the solvents indicated
and chemical shifts (d) were reported in ppm relative to an
internal standard (tetramethylsilane). Elemental analyses (C,
H, N), performed by a Carlo Erba 1106 elemental analyser,
resulted within 0.4% of theoretical values. UV spectra were
performed in ethanol using a Jasco V-520 (Tokyo) spectro-
photometer. The purity of compounds was checked by TLC
(pre-coated plates, silica gel Kieselgel 60 F254 (Merck) and
DC-Alufolien Aluminium oxide 60 F254 neutral, Typ.E
(Merck)).
3: H-NMR (CDCl3): 1.00 (t, J = 7 Hz, 3H, –CH2CH3),
1.21–1.80 (m, 3H, –CH–CH2CH3), 3.10 (s, 1H, OH), 4.00–
4.28 (m, 2H, HOCH2CH–), 7.24–7.38 (m, C5–H, pyridine)
7.42 (br s, 1H, –NH–), 8.10 (dd, J = 7.8 and 2.2 Hz, 1H, C4–H
pyridine), 9.00 (d, J = 2.2 Hz, 1H, C2–H pyridine).
4:
1H-NMR
(CDCl3):
1.92–2.18
(m,
2H,
–CH2CH2CH2–), 3.15 (s, 1H, –OH), 3.40–3.65 (m, 2H,
OCH2CH2CH2NH), 4.26–4.38 (m, 2H, OCH2CH2CH2NH),
7.25–7.37 (m, C5–H, pyridine), 7.48 (br s, 1H, –NH–), 8.08
(dd, J = 7.9, 2.0 Hz, 1H, C4–H pyridine), 8.98 (d, J = 2.0 Hz,
1H, C2–H pyridine).
4.1.4. General procedure for the synthesis of 5a–q
(Table 2)
Suitable hydroxyalkylnicotinamides 1–4 (16 mmol), dicy-
clohexylcarbodiimide (16 mmol) and fine 4-dimethyl-
aminopyridine (1.5 mmol) were added to a solution of the
drug (in acidic form) (15 mmol) in anhydrous acetonitrile
(150 ml). The mixtures were filtered and filtrates were dried
in vacuo. The residues were dissolved in CHCl3 and washed
first with a diluted NaHCO3 solution, then with diluted HCl
and eventually with water. The chloroform solutions were
dried over anhydrous Na2SO4 and the solvent was evaporated
in vacuo. The oily residues yielded pure compounds 5g and
5k, which were used for the successive reaction without
further purification. In the case of compound 5c, the residue
was induced to crystallize by adding few drops of diethyl
ether. The resulting solid was recrystallized by ethanol/
hexane (1:3). In all other cases the oily residues were purified
by flash chromatography using CHCl3 as eluant, with the
exception of compound 5e where the eluant used was
CHCl3/MeOH (99:1).
4.1.2. Synthesis of N-(2-hydroxyethyl)-3-nicotinamide 1
This compound was synthesized as described [16].
4.1.3. Synthesis of N-(2-hydroxyalkyl)-3-nicotinamide 2–4
Mixtures of ethyl nicotinate (0.02 mmol) and appropriate
amino alcohols (0.02 mmol) were stirred at 50 °C for 30 min,
and then left stirring at room temperature for 20–25 days.
The reaction mixtures were dried in vacuo to give com-
pounds [2–4] in quantitative yields. The crude oily products
were employed as such since they showed single spots by
TLC (CHCl3/MeOH 95:5).
1
2: H-NMR (CDCl3): 1.22 (d, J = 6 Hz, 3H, –CH3),
3.10–3.35 (br s, 1H, –OH), 3.55–3.75 (m, 1H, –CH–), 3.92–
4.05 (m, 1H, C1–H), 7.28–7.38 (m, C5–H, pyridine), 7.45 (br
s, 1H, –NH–), 8.10 (dd, 8.1 and 2.1 Hz, 1H, C4–H, pyridine),
8.60 (dd, J = 4.8, 1.6 Hz, 1H, C6–H pyridine), 9.00 (d,
J = 2.0 Hz, 1H, C2–H pyridine).

Table 2
Physical and chemical data of compounds 5a-q
D-COO-(chain)-NHCO-
5 a-q
N
Compound
Drug
DIK
Chain
m.p. °C
(solvent)
Yield (%)
45
Formula (MW)
1H NMR d
5a
CH2CH2
92–94 (EtAc)
C
22H19N3O3Cl2
3.72–3.82 (m, 2H, CH2NH), 3.85 (s, 2H, CH2COO), 4.37 (t, J = 7.0 Hz, 2H, CH2COOCH2), 6.41–7.38
(m, 10H, 7 aromatic + 2NH + C5–H pyridine), 7.88–8.05 (m, 1H, C4–H pyridine), 8.70 (dd,
J = 1.5, 5.2 Hz, 1H, C6–H, pyridine), 8.90 (d, J = 2.3 Hz, 1H, C2–H, pyridine) (CDCl3)
0.79 (d, J = 9.0 Hz, 6H, CH(CH3)2, 1.32 (d, J = 10.0 Hz, 3H, CH(CH3), 1.59–1.80 (m, 1H, CH(CH3)2),
2.32 (d, J = 10.0 Hz, 2H, CHCH2), 3.48–3.73 (m, 2H, CH2CH2), 3.75 (q, J = 10.0 Hz, 1H, CH(CH3),
4.02–4.30 (m, 2H, CH2CH2), 7.00, 7.15 (A2B2 system, J = 10.0 Hz, 4H, C6H4), 7.50–7.62 (m, 1H,
C5-H pyridine), 8.18 (dd, 1H, C4-H pyridine), 8.65–8.88 (m, 1H, C6–H pyridine), 9.00 (br s, 1H,
C2–H pyridine) (DMSO–d6)
(444.3)
5b
5c
IBU
CH2CH2
CH2CH2
72–75 (EtAc)
80
68
C
21H26N2O3
(354.5)
TOLM
115–118
(EtOH/hexane;
1:3)
C
23H23N3O4
2.35 (s, 3H, C6H4–CH3), 3.45–3.60 (m, 2H, CH2CH2), 3.82 (s, 3H, NCH3), 3.87 (s, 2H, CH2COO),
4.20–4.35 (m, 2H, CH2CH2), 6.10, 6.53 (AB system, J = 8 Hz, 2H, pyrrole), 7.25, 7.58 (A2B2 system,
J = 12.0 Hz, 4H, C6H4), 7.45–7.50 (m, 1H, C5–H pyridine), 8.13–8.25 (m, 1H, C4–H pyridine),
8.67–8.75 (m, 1H, C6–H pyridine), 8.80 (br s, 1H, NH), 8.95 (d, 1H, C2–H pyridine) (DMSO-d6)
1.55 (d, J = 10.0 Hz, 3H, CH3), 3.60–3.78 (m, 2H, CH2), 3.83 (q, 1H, CH), 4.20–4.46 (m, 2H, CH2), 6.70
(br s, 1H, NH), 7.20–7.60 (m, 9H, C6H4 + C6H5), 7.70–7.80 (m, 1H, C5–H pyridine), 7.90 (dd, 1H, C4–H
pyridine), 8.62 (d, 1H, C6–H pyridine), 8.81 (s, 1H, C2–H pyridine) (CDCl3)
(405.5)
5d
5e
5f
KET
TIAP
DIK
IBU
CH2CH2
Oil
Oil
Oil
Oil
77
C
24H22N2O4
(402.4)
CH2CH2
67
C
22H20N2O4S
1.66 (d, J = 7.0 Hz, 3H, CH3), 3.63–3.87 (m, 2H, CH2), 4.10 (q, J = 7.0 Hz, 1H, CH), 4.26–4.48 (m, 2H,
CH2), 6.45 (br s, 1H, NH), 6.98–7.90 (m, 8H, 7 aromatic + C5–H pyridine), 8.02 (d, 1H, C4–H pyridine),
8.70 (d, 1H, C6–H pyridine), 8.92 (s, 1H, C2–H pyridine) (CDCl3)
(408.5)
CH(CH3)CH2
CH(CH3)CH2
45
C
23H21N3O3Cl2
0.99 (d, J = 6.2 Hz, CH3), 2.99 (m, 2H, CH2), 3.80 (s, 2H, CH2COO), 4.73–4.86 (m, 1H, CH), 6.15–7.48
(m, 10H, 7 aromatic + NHCH2 + C5–H and C4–H pyridine), 8.25–8.38 (m, 2H, C6–H and C2–H
pyridine) (DMSO-d6)
(458.3)
5g
Quantitative
C22H28N2O3
0.8 (d, J = 6.6 Hz, 6H (CH3)2CH), 1.17–1.30 (m, 3H, CH3CHO), 1.45 (d, J = 7.2 Hz, 3H, CH3CHCO),
1.55–1.82 (m, 1H, (CH3)2CHCH2), 2.29–2.58 (m, 2H, (CH3)2CHCH2), 3.25–3.73 (m + q super-
imposed, 3H, CHCH2NH and CH3CHCOO), 5.0–5.2 (m, 1H, OCHCH2), 6.02
(368.5)
(br s, 1H, NH), 6.90–7.35 (m, 5H aromatic + C5–H pyridine), 7.70–7.80 (m, 1H, C4–H pyridine),
8.63–8.72 (m, 1H, C6–H pyridine), 8.83 (dd, J = 2.1 Hz, 1H, C2–H pyridine) (CDCl3)
1.32 (d, J = 7.6 Hz, 3H, CH3CHCH2), 1.58 (d, J = 7.1 Hz, 3H, CH3CHC6H4), 3.35–3.75 (m, 2H,
CH2NH), 3.80 (q, J = 7.1 Hz, 1H, CH3CHC6H4), 5.05–5.20 (m, 1H, CH3CHCH2), 6.48 (br s, 1H, NH),
7.20–7.89 (m, 11H, 9 aromatic + C5–H pyridine + C4–H pyridine), 8.53–8.77 (m, 2H, C6–H and
C2–H pyridine) (CDCl3)
5h
5i
KET
TIAP
DIK
CH(CH3)CH2
CH(CH3)CH2
Oil
Oil
71
48
62
C
25H24N2O4
(416.5)
C
23H22N2O4S
1.52–1.85 (2d superimposed, 6H, CH3CHCO + OCH(CH3)CH2), 3.20–3.45 (m, 2H, CHCH2NH),
3.48–3.67 (m, 1H, CH3CHCO), 5.21–5.38 (m, 1H, CH3CHCH2), 6.65 (br s, 1H, NHCO), 6.95–7.85
(m, 8H, 7 aromatic + C5–H pyridine), 7.86–7.91 (m, 1H, C4–H pyridine), 8.62–8.70 (m, 1H, C6–H
pyridine), 9.01 (s, 1H, C2–H pyridine) (DMSO-d6)
(422.5)
5j
CH2CH(CH2CH3) Oil
C
24H23N3O3Cl2
0.80–1.10 (m, 3H, CH3CH2), 1.44–1.78 (m, 2H, CH3CH2), 3.85 (s, 2H, CH2COO), 4.18–4.49 (m, 3H,
OCH2CH + CH2CHNH), 6.25–7.45 (m, 8H, 7 aromatic + C5–H pyridine), 7.93 (d, J = 7.2 Hz, 1H,
C4–H pyridine), 8.65 (s, 1H, C6–H pyridine), 8.9 (s, 1H, C2–H pyridine) (CDCl3)
(471.4)
(continued on next page)

Table 2
(continued)
Compound
Drug
IBU
Chain
m.p. °C
(solvent)
Yield (%)
67
Formula (MW)
1H NMR d
5k
CH2CH(CH2CH3) Oil
C
23H30N2O3
0.80–1.00 (t + 2d superimposed, 9H, (CH3)2CH + CH3CH2), 1.38–1.53 (m, 3H, CH3CHCO), 1.70–1.90
(m, 3H, (CH3)2CH + CH2CH3)), 2.22–2.45 (m, 2H, CHCH2C6H4), 3.68 (q, J = 7.2 Hz, 1H, CHC6H4),
4.02–4.34 (m, 3H, OCH2CHNH), 6.40 (t, J = 8.2 Hz, 1H, NHCO), 6.90–7.18 (m, 4H, aromatic),
7.26–7.33 (m, 1H, C5–H pyridine), 7.86–7.96 (m, 1H, C4–H pyridine), 8.64 (dd, J = 6.5 and 1.7 Hz,
1H, C6–H pyridine), 8.79–8.84 (m, 1H, C2–H pyridine) (CDCl3)
(382.5)
5l
KET
TIAP
DIK
IBU
CH2CH(CH2CH3) Oil
CH2CH(CH2CH3) Oil
70
72
50
77
C
26H26N2O4
0.89–1.00 (m, 3H, CH3CH2), 1.41–1.70 (m, 5H, CH3CH2 + CH3CH), 3.95 (q, J = 5.5 Hz, 1H, CHNH),
4.18–4.45 (m, 3H, CHCOOCH2), 6.43 (br s, 1H, NH), 7.30–7.80 (m, 10H, 9 aromatic + C5–H
pyridine), 7.90–8.10 (m, 1H, C4–H pyridine), 8.70 (br s, 1H, C6–H pyridine), 8.90 (br s, 1H,
C2–H pyridine) (CDCl3)
(430.5)
5m
5n
5o
C
24H24N2O4S
0.75–0.92 (m, 3H, CH3CH2), 1.35–1.80 (m, 5H, CH3CH + CH3CH2), 3.35–3.50 (m, 1H, CHCH3),
4.10–4.30 (m, 1H, CHNH), 6.80 (br s, 1H, NH), 7.10–7.90 (m, 7H, aromatic), 8.10–8.20 (m, 1H,
C5–H pyridine), 8.70–8.90 (m, 1H, C4–H pyridine), 9.00 (s, 1H, C6–H pyridine), 9.10–9.27 (m, 1H,
C2–H pyridine) (DMSO-d6)
(436.5)
CH2CH2CH2
CH2CH2CH2
Oil
Oil
C
23H21N3O3Cl2
1.90–2.10 (m, 2H, CH2CH2CH2), 3.47 (q, J = 6.3 Hz, 2H, CH2CH2CH2NH), 3.85 (s, 2H, CH2COO),
4.31 (t, J = 6.0 Hz, 2H, COOCH2CH2CH2), 6.48–7.40 (m, 9H, 7 aromatic + C5–H pyridine, NH), 8.1
(dd, J = 7.9 and 1.9 Hz, 1H, C4–H pyridine), 8.7 (dd, J = 4.8 and 1.5 Hz, 1H, C6–H pyridine), 9.0
(s, 1H, C2–H pyridine) (CDCl3)
(458.3)
C
22H28N2O3
0.85 (d, J = 6.6 Hz, 6H, CH(CH3)2), 1.47 (d, J = 7.2 Hz, 3H, CH3CHCO), 1.70–1.97 (m, 3H,
CH2CH2CH2NH + CH(CH3)2), 2.40 (d, J = 7.2 Hz, 2H, CH2C6H4), 3.18–3.42 (m, 2H,
CH2CH2CH2NH), 4.15 (t, J = 6.4 Hz, 2H, CH2CH2CH2NH), 6.78 (br s, 1H, NH), 6.98–7.38 (m, 5H,
4 aromatic + C5–H pyridine), 8.00–8.08 (m, 1H, C4–H pyridine), 8.65 (dd, J = 4.8 and 1.7 Hz, 1H,
C6–H pyridine), 8.85 (d, J = 1.8 Hz, 1H, C2–H pyridine) (CDCl3)
(368.5)
5p
5q
KET
CH2CH2CH2
CH2CH2CH2
Oil
Oil
65
30
C
25H24N2O4
1.50 (d, J = 7.1 Hz, 3H, CH3), 1.80-2.00 (m, 2H, CH2CH2CH2), 3.20-3.48 (m, 2H, CH2CH2CH2NH),
3.75-3.85 (d, J = 7.1 Hz, 1H, CH), 4.07-4.30 (m, 2H, CH2CH2CH2NH), 7.22-7.80 (m, 10H, 9 aromatic
+ C5-H pyridine), 8.05-8.12 (m, 1H, C4-H pyridine), 8.60 (dd, J = 5.2 and 2.3 Hz, 1H, C6-H), 8.95 (d,
J = 1.1 Hz, 1H, C2-H pyridine) (CDCl3)
(416.5)
TIAP
C
23H22N2O4S
1.60 (d, J = 7.2 Hz, 3H, CH3), 1.85–2.05 (m, 2H, CH2CH2CH2), 2.08–2.40 (br s, 1H, NH), 3.40–3.60
(m, 2H, CH2CH2CH2NH), 4.00–4.15 (m, 1H, CH), 4.18–4.30 (m, 2H, CH2CH2CH2NH), 6.80–7.85
(m, 8H, 7 aromatic + C5–H pyridine), 8.10 (dd, J = 8.2 and 2.2 Hz, 1H, C4–H pyridine), 8.72
(d, J = 4.4 Hz, 1H, C6–H pyridine), 9.00 (s, 1H, C2–H pyridine) (CDCl3)
(422.5)

722
L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
Rm0 values (linear regression) and calculated log P. PLS is a
chemiometric tool for extracting and rationalizing the infor-
mation from any multivariate description of a biological
system. Complexity reduction and data simplification are
two of the most important features of such a tool. PLS
condenses the overall information into two smaller matrixes,
namely the score plot (which shows the pattern of com-
pounds) and the loading plot (which shows the pattern of
descriptors). Since the chemical interpretation of score and
loading plots is simple and straightforward, PLS is usually
preferred to other non-linear methods, especially when the
noise is relatively high.
4.1.5. General procedure for the synthesis of 6a–q
(Table 3)
Solutions of (5a–q) (1 mmol) and CH3I (3.3 mmol) in
CH2Cl2/MeCN (3:1) (10 ml) were stirred in a closed system
at room temperature for 1 day for compounds (6a), (6d) and
(6e) and 6 days for all the others. The reaction mixtures were
brought down to dryness in vacuo. The oily residues, if pure
by TLC, were used in the successive reaction without further
purification, otherwise were purified by flash chromatogra-
phy, using CHCl3 as eluant for compounds (6e) and (6l) or
induced to crystallize with ethyl diether for (6b) or MeOH
(6c).
Score and loading plots are interconnected so that any
descriptor change in the loading plot is reflected by changes
in the position of compounds in the score plot. Pair wise
comparison can be made directly with interactive plots as
developed in the VolSurf program [14], and the relative
contributions to the property under study are shown in the
related descriptors space.
4.1.6. General procedure for the synthesis of 7a–q
(Table 4)
NaHCO3 (10 mmol) and Na2S2O4 (2.2 mmol) in an aque-
ous solution (150 ml) and ethyl acetate (30 ml) were added to
an ice-cooled solution of methyl iodide pyridinium deriva-
tives (6a–q) in degassed water (150 ml) under N2 atmo-
sphere. The reaction mixtures were stirred for 15 min at 0 °C
under N2, maintaining the pH at approximately 7 by addition
of NaHCO3. The organic phases were washed with water
several times, dried over Na2SO4, filtered and brought to
dryness in vacuo to afford 1,4-dihydropyridine derivatives
7a–q.
4.2.4. Predicting BBB permeation study using VolSurf
This study was conducted by a new predictive computer-
ized method [13] using descriptors derived from 3D molecu-
lar fields in estimating the BBB permeation of a large set of
compounds. The computational approach consists of four
main steps and implies the VolSurf procedure able to com-
press the relevant information present in 3D maps into a few
descriptors, simple to handle, to interpret, and correlate them
with experimental permeation. Three-dimensional molecular
fields in general contain large amount of data some of which
are redundant or not relevant for a given problem. Therefore,
novel tools are required to extract useful descriptors from
images of 3D molecular interactions fields (MIFs) and to link
experimental observations with molecular structures.
VolSurf automatically converts 3D molecular fields into
physico-chemically relevant molecular descriptors. In the
standard procedure, interaction fields with a water probe and
a hydrophobic probe are calculated for all the molecules in
the data set. However, grid maps produced by other probes
(ionic probes etc,) or by various molecular mechanics or
semi empirical approaches (e.g., electrostatic potential) can
also be used. Molecular recognition is achieved using image
analysis software coupled with external chemical knowl-
edge. Within this context, VolSurf selects the most appropri-
ate descriptors and parameterization according to the type of
3D maps under study.
4.2. Lipophilicity measurements
4.2.1. Rm determination
Rm values were determined by TLC using pre-coated glass
plates, C8 Whatman, 200 µm thickness and 20 µm particle
size with fluorescent indicator. The compounds were dis-
solved in acetone and the solutions were applied along a line
at about 2.5 cm from the bottom of the plate. The mobile
phases consisted of a mixture (200 ml) of various concentra-
tions of phosphate buffer (pH 7) and acetone. The phosphate
buffer was prepared according to Farmacopea Ufficiale della
Repubblica italiana XI ed. (F.U. XI) and then was diluted
1:10 (0.0067 M) with water. The mobile phase was allowed
to run 10 cm from the origin spots. The developed plates were
dried and detected by UV light.
Rm values vs. acetone concentrations were plotted and the
theoretical values of Rm at 0% of acetone in the mobile phase
were calculated by the least-squares method.
The molecular descriptors obtained refer to molecular size
and shape, to size and shape of both hydrophilic and hydro-
phobic regions and also to the balance between them [13].
4.2.2. Log P computation
Log P was estimated using the Kowwin computer pro-
gram, which applies atom/fragment contribution as de-
scribed [19].
Acknowledgements
4.2.3. Partial least square (PLS) discriminant analysis
The multivariate PLS projections to latent structure
method [20] was used to delineate the relationship between
We thank Professor Gabriele Cruciani for helpful discus-
sion on the VolSurf procedure.

Table 3
Physical and chemical data of compounds 6a–q
D-COO-(chain)-NHCO-
I-
N
+
6 a-q
CH3
Compound
Drug
DIK
Chain
m.p. °C
(solvent)
Oil
Yield (%)
88
Formula (MW)
1H NMR d (CDCl3)
6a
CH2CH2
C
23H22N3O3Cl2I
3.70–3.85 (m, 2H, CH2CH2NH), 3.92 (s, 2H, CH2CO), 4.35–4.44 (m, 2H, CH2CH2NH), 4.48 (s, 3H,
N+CH3), 6.38–7.35 (m, 8H, 7 aromatic + NH), 7.90–8.00 (m, 1H, C5–H pyridinium), 8.60 (br s, 1H,
C4–H pyridinium), 8.85–9.00 (m, 1H, C6–H pyridinium) (CDCl3)
(586.3)
6b
IBU
CH2CH2
144–6 (EtOH) 90
C
22H29N2O3I
0.80 (d, J = 9 Hz, 6H, CH(CH3)2), 1.32 (d, J = 10 Hz, 3H, CHCH3), 1.60–1.88 (m, 1H, CH(CH3)2),
2.35 (d, J = 10 Hz, CH2C6H4), 3.43–3.80 (m; 3H, CH2CH2NH + C6H4CHCH3), 4.02–4.28 (m, 2H,
OCH2CH2), 4.40 (s, 3H, N+CH3), 7.05, 7.15 (AB system, 4H, C6H4), 8.19–8.31 (m, 1H, C5–H
pyridinium), 8.82 (d, J = 7.7 Hz, 1H, C4–H pyridinium), 9.05–9.28 (m, 2H, NH + C6–H pyridinium),
9.36 (s, 1H, C2–H pyridinium) (DMSO-d6)
(496.4)
6c
6d
6e
6f
TOLM
KET
CH2CH2
107–111
(MeOH)
79
52
55
65
C
24H26N3O4I
2.35 (s, 3H, C6H4CH3), 3.40–3.65 (m, 2H, CH2CH2NH), 3.75 (s, 3H, N+CH3), 3.85 (s, 2H, CH2COO),
4.18–4.30 (m, 2H, CH2CH2NH), 6.05, 6.50 (AB system, J = 8.0 Hz, 2H, pyrrole), 7.25, 7.58 (AB
system, J = 7.7 Hz, 4H, C6H4) 8.10 (m, 1H, C4–H pyridinium), 8.60 (m, 1H, C6–H pyridinium), 8.85
(s, 1H, NH), 8.98 (s, 1H, C2–H pyridinium) (DMSO-d6)
(547.4)
CH2CH2
Oil
C
25H25N2O4I
1.55 (d, J = 6.0 Hz, 3H, CHCH3), 3.60–3.80 (m, 3H, CH2CH2NH + NH), 3.98 (q, 1H, CHCH3),
4.25–4.45 (m, 2H, CH2CH2NH), 4.60 (s, 3H, N+CH3), 7.33–7.80 (m, 9H, aromatic), 7.90–8.08 (m, 1H,
C5–H pyridinium), 8.78 (d, J = 7.2 Hz, 1H, C4–H pyridinium), 8.85–9.00 (m, 1H, C6–H pyridinium),
9.99 (s, 1H, C2–H pyridinium) (CDCl3)
(544.4)
TIAP
DIK
CH2CH2
Oil
C
23H23N2O4SI
1.50 (d, J = 7.1 Hz, 3H, CH3), 3.45–3.75 (m, 2H, CH2CH2NH), 4.18–4.45 (m and s superimposed, 6H,
CH2CH2NH + CHCH3 + N+CH3), 6.85 (s, 1H, NH), 6.90–8.35 (m, 7H, aromatic), 8.70–8.90 (m, 1H,
C5–H pyridinium), 9.10 (d, J = 6.0 Hz, 1H, C4–H pyridinium), 9.15–9.30 (m, 1H, C6–H pyridinium),
9.38 (s, 1H, C2–H pyridinium) (DMSO-d6)
(550.4)
CH(CH3)CH2
151–3 (EtAc)
C
24H24N3O3Cl2I
1.27 (d, J = 6.3 Hz, 3H, CH3), 3.40–3.70 (m, 2H, CH2NH), 3.80 (s, 2H, C6H4CH2CO), 4.38 (s, 3H,
N+CH3), 4.99–5.19 (m, 1H, CH), 6.11–7.22 (m, 5H, aromatic), 7.49 (d, J = 8.0 Hz, 2H, C3–H and
C5–H, dichlorophenyl), 8.16 (t, J = 7.3, 1H, C5–H pyridinium), 8.8 (d, J = 8.1 Hz, 1H, C4–H
pyridinium), 9.07 (d, J = 5.8 Hz, 1H, C6–H pyridinium), 9.13–9.24 (m, 1H, C2–H pyridinium), 9.32
(s, 1H, C6H3–NH–C6H4) (DMSO-d6)
(600.3)
6g
IBU
CH(CH3)CH2
Oil
89
C
23H31N2O3I
0.70–0.85 (m, 6H, CH(CH3)2), 1.08–1.38 (m, 6H, CH3CHO + CH3CHCO), 1.50–1.80 (m, 1H,
CH(CH3)2), 2.20–2.37 (m, 2H, CH2N), 3.60 (q, J = 7.0 Hz, 1H, CH3CH), 4.39 (s, 3H, N+CH3),
4.90–5.07 (m, 1H, CHCH3), 6.88–7.27 (m, 4H, aromatic), 8.10–8.28 (m, 1H, C5–H piridinium),
8.65–8.90 (m, 1H, C4–H pyridinium), 9.01–9.20 (m, 2H, C6–H pyridinium + NH), 9.25–9.35 (m, 1H,
C2–H pyridinium) (DMSO-d6)
(510.4)
6h
6i
KET
CH(CH3)CH2
CH(CH3)CH2
Oil
Oil
80
68
C
26H27N2O4I
1.20 (d, J = 6.3 Hz, 3H, CH3CH), 1.40 (d, J = 6.8 Hz, 3H, CH(CH3)), 3.15 (d, J = 5.2, 2H, CH2NH), 3.85
(q, J = 6.3 Hz, 1H, CH3CH), 4.10 (q, J = 6.8 Hz, 1H, CHCH3), 4.38 (s, 3H, N+CH3), 7.33–7.73 (m, 9H,
aromatic), 8.12–8.27 (m, 1H, C5–H piridinium), 8.64 (d, J = 7.9 Hz, 1H, C4–H pyridinium), 8.98–9.10
(m, 2H, C6–H pyridinium + NHCO), 9.25 (s, 1H, C2–H, pyridinium) (DMSO-d6)
1.13–1.40 (m, 6H, CH3CHO + CH3CHCH2), 3.10–3.60 (m, 3H, CH2NH + CH3CHCO), 4.40 (s, 3H,
N+CH3), 4.89-5.22 (m, 1H, CH3CHCH2), 6.50–7.85 (m, 7H, C6H4 + thiophene), 8.00–8.30 (m, 1H,
C5–H pyridinium), 8.60–9.45 (m, 3H, C4–H, C6–H, C2–H pyridinium) (DMSO-d6)
(558.4)
TIAP
C
24H25N2O4SI
(564.4)
(continued on next page)

Table 3
(continued)
Compound
Drug
DIK
Chain
m.p. °C
(solvent)
Yield (%)
95
Formula (MW)
1H NMR d
6j
CH2CH(CH2CH3) Oil
C
25H26N3O3ClI
0.83–1.12 (m, 3H, CH3CH2), 1.40–1.75 (m, 2H, CH3CH2), 3.82 (s, 2H, CH2COO), 4.10-4.45 (m,
CH2CHNH + CH2CHNH), 4.47 (s, 3H, N+CH3), 6.25–7.42 (m, 7H, 6 aromatic + C5–H pyridinium),
7.85–8.00 (m, 1H, C4–H pyridinium), 8.55–8.75 (m, 1H, C6–H pyridinium), 8.90 (s, 1H, C2–H pyri-
dinium) (CDCl3)
(614.2)
6k
IBU
CH2CH(CH2CH3) Oil
Quantitative
C
24H33N2O3I
0.80–1.05 (m, 9H, CH(CH3)2 + CH3CH2), 1.45–1.53 (m, 3H, CH3CHCO), 1.60–1.90 (m, 3H,
(CH3)2CH + CHCH2CH3), 2.33, 2.37 (2d, J = 7.09 Hz, J = 7.14 Hz, 2H, CH2C6H4), 2.39–2.48 (m,
1H, CHNH), 3.68–3.92 (m, 1H, CHCO), 4.20–4.40 (m, 2H, OCH2), 4.55 (s, 3H, N+CH3), 6.95-7.35
(m, 4H, aromatic), 7.94–8.08 (m, 1H, C5–H pyridiniuim), 8.40–8.53 (m, 1H, C4–H pyridinium), 8.78
(br s, 1H, NH), 8.85–8.93 (m, 1H, C6–H pyridinium), 9.90 (s, 1H, C2–H pyridinium) (CDCl3)
0.75–0.95 (m, 3H, CH2CH3), 1.30–1.70 (m, 5H CHCH3 + CH2CH3), 3.85–4.00 (s, 1H, CHCH3),
4.03–4.22 (m, 2H, OCH2), 4.40 (s, 3H, N+CH3), 7.30–7.80 (m, 9H, aromatic), 8.10–8.25 (m, 1H,
C5–H pyridinium), 8.70–8.82 (m, 2H, NH + C4–H pyridinium), 9.00–9.12 (m, 1H, C6–H pyridinium),
9.25 (s, 1H, C2–H pyridinium) (DMSO-d6)
(524.2)
6l
KET
TIAP
DIK
CH2CH(CH2CH3) Oil
CH2CH(CH2CH3) Oil
45
62
80
C
27H29N2O4I
(572.1)
6m
6n
C
25H27N2O4SI
0.78–0.95 (m, 3H, CH3CH2), 1.35–1.75 (m, 5H, CH3CH + CH3CH2), 3.40–3.50 (m, 1H, CHCH3),
4.10–4.32 (m, 1H, CHNH); 4.40 (s, 3H, N+CH3), 7.00 (br s, 1H, NH), 7.12–7.85 (m, 7H, aromatic),
8.10–8.18 (m, 1H, C5–H pyridinium), 8.70–8.90 (m, 1H, C4–H pyridinium), 9.10 (s, 1H, C6–H
pyridinium), 9.26–9.40 (m, 1H, C2–H pyridinium) (DMSO-d6)
(578.1)
CH2CH2CH2
CH2CH2CH2
Oil
Oil
C
24H24N3O3Cl2I
1.79–1.98 (m, 2H, CH2CH2CH2NH), 3.25–3.50 (q, J = 6.4 Hz, 2H, CH2CH2CH2NH), 3.80 (s, 2H,
CH2COO), 4.10–4.20 (t, J = 6.3 Hz, 2H, CH2CH2CH2NH), 4.38 (s, 2H, N+CH3), 6.25 (d, J = 7.9, 1H,
NH), 6.80–7.55 (m, 7H, aromatic), 8.20–8.35 (m, 1H, C5-H pyridinium), 8.85 (d, J = 8.2 Hz, 1H,
C4–H pyridinium), 9.05 (d, J = 5.1 Hz, 1H, C6–H pyridinium), 9.33 (s, 1H, C2-H pyridinium)
(DMSO-d6)
(559.0)
6o
IBU
Quantitative
C23H31N2O3I
1.85 (d, J = 6.6 Hz, 3H, CH(CH3)2), 1.45 (d, J = 7.2 Hz, 3H, CHCH3), 1.60–1.90 (m, 1H, CH(CH3)2),
1.95–2.10 (m, 2H, CH2CH2CH2NH), 2.40 (d, J = 7.1 Hz, 2H, CH2C6H4), 3.45–3.57 (m, 2H,
CH2CH2CH2NH), 3.75 (q, J = 7.3 Hz, 1H, CHCH3), 4.00–4.30 (m, 2H, CH2CH2CH2NH), 4.60 (s, 3H,
N+CH3), 7.00–7.25 (m, 4H, aromatics), 8.01–8.10 (m, 1H, C5–H pyridinium), 8.65–8.80 (m, 2H, NH
+ C4–H pyridinium), 8.95 (d, J = 5.1 Hz, 1H, C6–H pyridinium), 10.15 (s, 1H, C2–H pyridinium)
(DMSO-d6)
(510.1)
6p
6q
KET
CH2CH2CH2
CH2CH2CH2
Oil
Oil
68
80
C
24H25N2O4I
1.40 (d, J = 7.13 Hz, 3H, CHCH3), 1.75–1.90 (m, 2H, CH2CH2CH2NH), 3.09–3.18 (m, 2H,
CH2CH2CH2NH), 3.85–4.00 (q, 1H, CH3CH), 4.05–4.20 (m, 2H, CH2CH2CH2NH), 4.40 (s, 3H,
N+CH3), 7.40–7.80 (m, 9H, aromatic), 8.18-8.28 (m, 1H, C5-H pyridinium), 8.85 (d, J = 8.2 Hz, 1H,
C4–H pyridinium), 8.90–9.09 (m + s superimposed, 2H, C6–H pyridinium + NHCO), 9.22 (s, 1H,
C2-H pyridinium) (DMSO-d6)
(532.1)
TIAP
C
24H25N2O4SI
1.65–1.80 (m, 2H, CH2CH2CH2), 1.82–1.90 (m, 3H, CHCH3), 3.09–3.18 (m, 2H, CH2CH2CH2NH),
4.02 (q, J = 8.1 Hz, CHCH3), 4.10–4.28 (m, 2H, CH2CH2CH2NH), 4.37 (s, 3H, N+CH3), 7.45–7.85
(m, 7H, aromatic), 8.08-8.18 (m, 1H, C5-H pyridinium), 8.85-8.95 (m, 1H, C4-H pyridinium),
9.04–9.12 (m and s superimposed, 2H, NH + C6–H pyridinium), 9.30 (s, 1H, C2–H pyridinium)
(DMSO-d6)
(564.1)

Table 4
Physical and chemical data of compounds 7a–q graphique
D-COO-(chain)-NHCO-
N
CH3
7 a - q
Compound
Drug
DIK
Chain
Yield (%)
Formula (MW)
23H23N3O3Cl2 2.80 (br s, 2H, CH2 dihydropyridine), 3.00 (s, 3H, NCH3), 3.72–3.85 (m, 2H, CH2NH), 3.95 (s, 2H, CH2COOCH2),
1H NMR d (CDCl3)
7a
CH2CH2
Quantitative
C
(459.1)
4.15–4.26 (m, 2H, CH2COOCH2), 4.58–4.75 (m, 1H, C5–H dihydropyridine), 5.20 (br s, 1H, NH), 5.55 (m, 1H,
C6–H dihydropyridine), 6.90 (s, 1H, C2–H dihydropyridine), 6.35–7.34 (m, 8H, 7 aromatic + NH) (CDCl3)
0.90 (d, J = 6.6 Hz, 6H, CH(CH3)2, 1.45 (d, J = 7.1 Hz, 3H, CHCH3), 1.75–1.95 (m, 1H CH(CH3)2), 2.45
(d, J = 7.5 Hz, 2H, CH2C6H4), 2.88 (br s, 2H CH2 dihydropyridine), 2.95 (s, 3H, NCH3), 3.47–3.58 (m, 2H,
CH2CH2NH), 3.68–3.85 (m, 1H, C6H4CHCH3), 4.08–4.27 (m, 2H, CH2CH2NH), 4.65–4.73 (m, 1H, C5–H dihydro-
pyridine), 5.23 (br s, 1H, NH), 5.65–5.70 (m, 1H, C6–H dihydropyridine), 6.95 (s, 1H, C2–H dihydropyridine),
7.04–7.20 (A2B2 system, 4H, C6H4) (CDCl3)
7b
IBU
CH2CH2
Quantitative
C22H30N2O3
(370.2)
7c
TOLM
KET
CH2CH2
CH2CH2
Quantitative
83
C
24H27N3O4
2.35 (s, 3H, C6H4CH3), 2.90 (br s, 2H, CH2 dihydropyridine), 3.40–3.65 (m, 2H, CH2CH2), 3.85 (s, 3H, NCH3), 3.90
(s, 2H, CH2COO), 4.22–4.38 (m, 2H, CH2CH2), 4.58–4.70 (m, 1H, C5–H dihydropyridine), 6.08, 6.50 (AB system,
J = 8 Hz, 2H, pyrrole), 7.00 (s, 1H, C2–H dihydropyridine), 7.23, 7.58 (A2B2 system, J = 12.0 Hz, 4H, aromatic),
8.95 (s, 1H, NH) (DMSO-d6)
(421.2)
7d
C
25H26N2O4
1.53 (d, J = 9.8 Hz, 3H, CH3), 2.90 (2s superimposed, 5H, NCH3 + CH2 dihydropyridine), 3.48–3.51 (m, 2H,
CH2CH2NH), 3.83 (q, J = 9.8 Hz, 1H, CHCH3), 4.08–4.20 (m, 2H, CH2CH2NH), 4.55–4.68 (m, 1H, C5–H dihydro-
pyridine), 5.28 (br s, 1H, NH), 5.57–5.68 (m, 1H, C6–H dihydropyridine), 6.95 (s, 1H, C2–H dihydropyridine),
7.35–7.85 (m, 9H, aromatic) (CDCl3)
(418.2)
7e
7f
TIAP
DIK
CH2CH2
52
57
C
23H24N2O4S
1.56 (d, J = 7.6 Hz, 3H, CHCH3), 2.83 (s, 3H, NCH3), 2.95 (br s, 2H, CH2 dihydropyridine), 3.50–3.70 (m, 2H,
CH2CH2NH), 4.18–4.39 (m, 2H, CH2CH2NH), 4.51–4.68 (m, 1H, C5–H dihydropyridine), 5.40–5.62 (m, 2H,
NH + C6–H dihydropyridine), 6.90 (br s, 1H, C2–H dihydropyridine), 7.23–7.81 (m, 7H, aromatic) (CDCl3)
(424.1)
CH(CH3)CH2
C24H25N3O3Cl2 0.85–1.02 (m, 3H, CHCH3), 2.85 (br s, 2H, CH2 dihydropyridine), 2.90 (s, 3H, NCH3), 3.52–3.80 (m, 2H, CH2NH),
(473.1)
4.43 (s, 2H, CH2COO), 4.52–4.68 (m, 1H, C5–H dihydropyridine), 5.10 (br s, 1H, NHC6H4), 5.20–5.45 (m, 1H,
OCHCH3), 5.50–5.66 (m, 1H, C6–H dihydropyridine), 6.80 (s, 1H, C2–H dihydropyridine), 6.90–7.48 (m, 7H,
aromatic), 10.10 (br s, 1H, NHCO) (CDCl3)
7g
IBU
CH(CH3)CH2
68
C
23H32N2O3
0.90 (d, J = 6.6 Hz, 6H, CH(CH3)2), 1.15–1.25 (m, 3H, OCHCH3), 1.40–1.50 (m, 3H, OCHCH3), 1.70–1.97 (m, 1H,
(CH3)2CH), 2.38–2.49 (d, J = 7.2 Hz, 2H, CH2C6H4), 2.65 (br s, H, CH2 dihydropyridine), 2.90 (s, 3H, NCH3),
3.20–3.75 (m, 3H, CH3CHCO + CH2NH), 4.50–4.70 (m, 1H, OCHCH3), 4.90–5.05 (m, 1H, C5–H dihydro-
pyridine), 5.57–5.70 (m, 1H, C6–H dihydropyridine), 6.90 (s, 1H, C2–H dihydropyridine), 7.00–7.23 (m, 4H,
aromatic) (CDCl3)
(384.2)
7h
7i
KET
TIAP
DIK
CH(CH3)CH2
60
78
50
C
26H28N2O4
1.25 (d, J = 8.08 Hz, 3H, CHCH3), 1.50 (d, 7.2 Hz. 3H, CHCH3), 2.66 (s, 2H, C4–H dihydropyridine), 2.86 (s, 3H,
NCH3), 3.23–3.29 (m, 2H, CH2NH), 3.70–3.85 (q, J = 8.1 Hz, 1H, CHCH3), 4.48–4.59 (m, 1H, C5–H dihydropyri-
dine), 4.90–5.15 (m, 2H, CHCH3 + NH), 6.88 (s, 1H, C2–H dihydropyridine), 7.31–7.77 (m, 9H, aromatic) (CDCl3)
1.30–1.50 (2d superimposed, 6H, OCHCH3 + CH3CHCO), 2.85 (s, 3H, NCH3), 2.92 (s, 2H, CH2 dihydropyridine),
3.55–3.80 (m, 1H, CH3CHCO), 4.50–4.72 (m, 1H, C5–H dihydropyridine), 5.05–5.25 (m, 1H, OCHCH3), 5.50–5.72
(m, 1H, C6–H dihydropyridine), 6.85 (s, 1H, C2–H dihydropyridine), 7.10–7.92 (m, 7H, aromatic) (CDCl3)
(432.2)
CH(CH3)CH2
C
24H26N2O4S
(438.2)
7j
CH2CH(CH2CH3)
C25H27N3O3Cl2 0.85–1.12 (m, 3H, CH3CH2), 1.48–1.80 (m, 2H, CH3CH2), 2.85 (s, 3H, NCH3), 2.92 (br s, 2H, CH2 dihydropyridine),
(487.1)
3.80 (s, 2H, C6H4CH2CO), 4.15–4.35 (m, 3H, OCH2CH + CH2CHNH), 4.45–4.68 (m, 1H, C5-H dihydropyri-
dine), 5.12 (br s, 1H, NHC6H4), 5.50–5.65 (m, 1H, C6–H dihydropyridine), 6.25–7.49 (m, 8H, 7 aromatic + C2–H
dihydropyridine), 9.95 (br s, 1H, NHCO) (CDCl3)
(continued on next page)

Table 4
(continued)
Compound
Drug
IBU
Chain
Yield (%)
67
Formula (MW)
24H33N2O3
(397.5)
1H NMR d (CDCl3)
7k
CH2CH(CH2CH3)
C
0.80–0.95 (m, 9H, CH(CH3)2 + CH2CH3), 1.40–1.54 (m, 3H, CH3CHCO), 1.60–2.00 (m, 3H, (CH3)2CH + CH2CH3),
2.40–2.50 (m, 2H, CH2C6H4), 2.90 (2 s superimposed, 5H, NCH3, CH2 dihydropyridine), 3.60–3.80 (s, 1H, CHCH3),
4.08–4.23 (m, 3H, CH2CHN), 4.82–5.05 (m, 1H, C5–H dihydropyridine), 5.62–5.70 (m, 1H, C6–H dihydropyri-
dine), 6.94 (s, 1H, C2–H dihydropyridine), 7.00–7.30 (m, 4H, aromatic) (CDCl3)
7l
KET
TIAP
DIK
CH2CH(CH2CH3)
CH2CH(CH2CH3)
CH2CH2CH2
82
47
78
65
68
72
C
27H30N2O4
0.89 (t, J = 7.3 Hz, 3H, CH2CH3), 1.40–1.53 (m and d superimposed, J = 7.2 Hz, 5H, CH2CH3 + CHCH3), 2.85 (s, 2H,
CH2 dihydropyridine), 2.90 (s, 2H, NCH3), 3.70–3.85 (q, J = 7.2 Hz, 1H, CHCH3), 3.98–4.15 (m, 1H, OCH2CH),
4.55–4.70 (m, 3H, OCH2CH + C5–H dihydropyridine), 5.00 (br s, 1H, NH), 5.50–5.55 (m, 1H, C6–H dihydro-
pyridine), 6.90 (s, 1H, C2–H dihydropyridine), 7.30–7.80 (m, 9H, aromatic) (CDCl3)
(446.5)
7m
7n
7o
7p
7q
C
25H28N2O4S
0.80–0.95 (m, 3H, CH3CH2), 1.38–1.91 (m, 5H, CH3CH + CH3CH2), 2.90 (s, 3H, NCH3), 2.96 (m, 2H, CH2
dihydropyridine), 3.41–3.55 (m, 1H, CHCH3), 4.05–4.25 (m, 1H, CHNH), 4.57–4.72 (m, 1H, C5–H dihydropyri-
dine), 5.40–5.58 (m, 1H, C6–H dihydropyridine), 6.75 (br s, 1H, NH), 6.95 (br s, 1H, C2–H dihydropyridine),
7.15–8.00 (m, 7H, aromatic) (CDCl3)
(452.6)
C
24H25N3O3Cl2 1.79–2.05 (m, 2H, CH2CH2CH2), 2.85 (br s, 2H, CH2 dihydropyridine), 2.92 (s, 3H, NCH3), 3.40–3.56 (m, 2H,
(474.4)
CH2CH2CH2NH), 3.91 (s, 2H, CH2COO), 4.28 (t, J = 6.3 Hz, 2H, CH2CH2CH2NH), 4.55–4.68 (m, 1H, C5–H
dihydropyridine), 5.60–5.72 (m, 1H, C6-H dihydropyridine), 6.45–7.38 (m, 9H, 7 aromatic + C2-H dihydropyri-
dine + NH) (CDCl3)
IBU
CH2CH2CH2
C
23H32N2O3
0.80 (d, J = 7.0 Hz, 6H, CH(CH3)2, 1.45 (d, J = 7.2 Hz, 3H, CH3CHO), 1.65–1.88 (m, 3H, CH2CH2CH2NH +
CH(CH3)2, 2.45 (d, J = 7.2 Hz, 2H, CH2C6H4), 2.90 (br s, 2H, CH2 dihydropyridine), 2.96 (s, 3H, NCH3), 3.16–3.38
(m, 2H, CH2CH2CH2NH), 4.10 (t, J = 6.4 Hz, 2H CH2CH2CH2NH), 4.60–4.69 (m, 1H, C5–H dihydropyridine),
5.60–5.68 (m, 1H, C6–H dihydropyridine), 6.78–7.38 (m, 6H, 4 aromatic, NH, C2–H dihydropyridine) (CDCl3)
1.55 (d, J = 7.2 Hz, 3H, CH3CH), 1.67–1.90 (m, 2H, CH2CH2CH2NH), 2.90 (s, 3H, N+CH3), 3.10 (br s, 2H, CH2
dihydropyridine), 3.15–3.38 (m, 2H, CH2CH2CH2NH), 3.80 (q, J = 7.2 Hz, 1H, CHCH3), 4.00–4.20 (m, 2H,
CH2CH2CH2NH), 4.60–4.75 (m, 1H, C5–H dihydropyridine), 5.48 (br s, 1H, NHCO), 5.60–5.70 (m, 1H, C6–H
dihydropyridine), 6.95 (s, 1H, C2–H dihydropyridine), 7.35–7.84 (m, 9H, aromatic) (CDCl3)
1.85 (d, J = 6.4 Hz, 3H, CH3CH), 1.90–2.02 (m, 2H, CH2CH2CH2NH), 2.90 (s, 3H, NCH3), 3.10 (br s, 2H, CH2
dihydropyridine), 3.30–3.45 (m, 2H, CH2CH2CH2NH), 4.28–4.40 (m, 2H, CH2CH2CH2NH), 4.60–4.76 (m, 1H,
C5–H dihydropyridine), 5.55 (br s, 1H, NHCO), 5.60–5.70 (m, 1H, C6–H dihydropyridine), 6.98 (s, 1H, C2–H
dihydropyridine), 6.93–7.90 (m, 7H, aromatic) (CDCl3)
(384.5)
KET
TIAP
CH2CH2CH2
C
26H28N2O4
(432.5)
CH2CH2CH2
C
24H26N2O4S
(438.5)

L. Perioli et al. / European Journal of Medicinal Chemistry 39 (2004) 715–727
727
[12] M. Hull, K. Lieb, B.L. Fiebich, Expert Opin. Inv. Drug 9 (2000)
671–683.
References
[13] P. Crivori, G. Cruciani, P.A. Carrupt, B. Testa, J. Med. Chem. 43
(2000) 2204–2216.
[14] G. Cruciani, P. Crivori, P.A. Carrupt, B. Testa, J. Mol, Struc-
Theochem 503 (2000) 17–30.
[1] R. Diaz Brinton, R.S. Yamazaki, Pharm. Res. 15 (1998) 386–398.
[2] P. Eikelenboom, A.J.M. Rozemuller, J.J.M. Hoozemans, R. Veerhuis,
W.A. Van Gool, Alz. Dis. Assoc. Dis. 14 (2000) S54–S61.
[3] M.A.A. Blom, M.G.H. van Twillert, S.C. de Vries, F. Engels,
C.E. Finch, R. Veerhuis, P. Eikelenboom, Brain Res. 777 (1997)
210–218.
[15] E. Pop, Curr. Med. Chem. 4 (1997) 279–294; N. Bodor, in: E.B. Roch
(Ed.), Theory of Application of Bioreversible Carriers to Drug
Design, Pergamon, New York, 1985, pp. 95–120; E. Pop, E. Shek,
T. Murakami, N.S. Bodor, J. Pharm. Sci. 78 (1989) 609–616; E. Pop,
W. Whei-Mei, E. Shek, N. Bodor, J. Med. Chem. 32 (1989) 1774–
1781; E. Pop, W. Whei-Mei, N. Bodor, J. Med. Chem. 32 (1989)
1789–1795; E. Pop, W. Anderson, K. Prókai-Tátrai, M.E. Brewster,
M. Fregly, N. Bodor, J. Med. Chem. 33 (1990) 2216–2221;
M.J. Phelan, N. Bodor, Pharm. Res. 6 (1989) 667–676.
[16] T. Ogowa, K. Katayama, H. Maeda, Y. Kita, Chem. Pharm. Bull. 42
(1994) 1579–1589.
[4] K. Anderson, L.J. Launer, A. Ott, A.W. Hoes, M.M. Breteler, A. Hof-
man, Neurology 45 (1995) 1441–1445.
[5] W.F. Stewart, C. Kawas, M. Corrada, E.J. Metter, Neurology 48
(1997) 626–632.
[6] B.L. Flynn, K.A. Theesen, Ann. Pharmacother. 33 (1999) 840–849.
[7] K. Sugaya, T. Uz, V. Kumar, H. Manev, Jpn. J. Pharmacol. 82 (2000)
85–94.
[8] P.L. McGeer, M. Schulzer, E.G. McGeer, in: J. Rogers (Ed.), Neuroin-
flammatory Mechanism in Alzheimer’s Disease, Birkhaeuser Verlag,
Switzerland, Basel, 2001, pp. 53–64.
[17] C.B.C. Boyce, B.V. Milborrow, Nature 208 (1995) 537–539.
[18] G.L. Biagi, A.M. Barbaro, M.F. Gamba, M.C. Guerra, J. Chromatog.
41 (1969) 371–379.
[9] J.C.S. Breitner, Annu. Rev. Med. 47 (1996) 401–411.
[10] P.S. Aisen, K.L. Davis, Am. J. Psychiatr. 151 (1994) 1105–1113.
[11] P.S. Aisen, Drug Develop. Res. 56 (2001) 421–427.
[19] W.M. Meylan, P.H. Howard, J. Pharm. Sci. 84 (1995) 83–92.
[20] H. Wold, Research Papers in Statistics: Ferstschrift for J. Neyman,
Wiley, New York, 1966.