Synthesis and pharmacological properties of a new hydrophilic and orally bioavailable 5-HT4 antagonist

Article abstract of DOI:10.1016/j.ejmech.2013.03.060

5-HT₄ receptor antagonists have been suggested to have clinical potential in treatment of atrial fibrillation, diarrhea-prone irritable bowel syndrome and urinary incontinence. Recently, the use of 5-HT₄ antagonists has been suggeste

Full text of DOI:10.1016/j.ejmech.2013.03.060

European Journal of Medicinal Chemistry 64 (2013) 629e637
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and pharmacological properties of a new hydrophilic and
orally bioavailable 5-HT4 antagonist
,
,
,
,c
Bjarne Brudeli a d, Lise Román Moltzau b c, Cam H.T. Nguyen b c, Kjetil Wessel Andressen b
,
Nils Olav Nilsen a, Finn Olav Levy b c, Jo Klaveness a
a Drug Discovery Laboratory AS, Oslo Innovation Center, N-0349 Oslo, Norway
b Department of Pharmacology, Faculty of Medicine, University of Oslo and Oslo University Hospital, P.O. Box 1057, Blindern, N-0316 Oslo, Norway
c K.G. Jebsen Cardiac Research Centre and Center for Heart Failure Research, Faculty of Medicine, University of Oslo, N-0316 Oslo, Norway
d Department of Medicinal Chemistry, School of Pharmacy, University of Oslo, N-0316 Oslo, Norway
,
,d,
*
a r t i c l e i n f o
a b s t r a c t
Article history:
5-HT4 receptor antagonists have been suggested to have clinical potential in treatment of atrial fibril-
lation, diarrhea-prone irritable bowel syndrome and urinary incontinence. Recently, the use of 5-HT4
antagonists has been suggested to have a therapeutic benefit in heart failure. Affinity for the hERG po-
tassium ion channel and increased risk for prolonged QT intervals and arrhythmias has been observed for
several 5-HT4 ligands. Serotonin may also have beneficial effects in the central nervous system (CNS)
through stimulation of the 5-HT4 receptor, and reduced distribution of 5-HT4 antagonists to the CNS may
therefore be an advantage. Replacing the amide and N-butyl side chain of the 5-HT4 receptor antagonist
SB207266 with an ester and a benzyl dimethyl acetic acid group led to compound 9; a hydrophilic 5-HT4
antagonist with excellent receptor binding and low affinity for the hERG potassium ion channel. To in-
crease oral bioavailability of carboxylic acid 9, two different prodrug approaches were applied. The tert-
butyl prodrug 11 did not improve bioavailability, and LC-MS analysis revealed unmetabolized prodrug in
the systemic circulation. The medoxomil ester prodrug 10 showed complete conversion and sufficient
bioavailability of 9 to advance into further preclinical testing for treatment of heart failure.
Ó 2013 Elsevier Masson SAS. All rights reserved.
Received 17 October 2012
Received in revised form
22 March 2013
Accepted 27 March 2013
Available online 9 April 2013
Keywords:
Bioavailability
hERG
5-HT4 antagonist
Hydrophilic
Pharmacokinetic
Prodrug
1. Introduction
failing hearts [4]. Further, increased expression of the 5-HT4 re-
ceptor may indicate a role of the serotonergic system in this con-
Serotonin (5-hydroxytryptamine; 5-HT) is a monoaminergic
neurotransmitter in the central and peripheral nervous system.
Today, 14 different human serotonin receptors are known, divided
into 7 distinct families (5-HT1e7) [1]. Pharmaceuticals acting on
serotonin receptors play important roles in several pathological
conditions. The first line treatment of migraine is triptans, a class of
selective serotonin agonists that affect the 5-HT1B and 5-HT1D re-
ceptors in cranial artery smooth muscle [2]. Nausea and emesis after
chemotherapy are treated with e.g. ondansetron and tropisetron,
selective serotonin antagonists that affect the 5-HT3 receptor in the
chemoreceptor trigger zone and the gastro-intestinal system.
Serotonin causes increased rate and force of contraction through
5-HT4 receptors in human atria [3]. Recent investigations have
revealed that the human cardiac ventricles show increased
response to serotonin through 5-HT4 receptors in infarcted and
dition. Recently, blocking of the 5-HT4 receptor has been
investigated in both experimental animals and humans, and the
use of 5-HT4 receptor antagonists may be a new therapeutic
intervention in treatment of congestive heart failure [5,6].
Irritable bowel syndrome (IBS) is a gastrointestinal disorder
characterized by abdominal pain, bloating and altered bowel
habits. An increased understanding of the role of serotonin in
regulation of GI motility, secretion and visceral sensitivity has led to
the development of 5-HT receptor modulators for treatment of IBS
[7]. Both the 5-HT3 receptor antagonists alosetron and cilansetron
and the 5-HT4 agonists cisapride and tegaserod have been devel-
oped for treatment of diarrhea- and constipation-predominant
forms of IBS, respectively. However, serious side effects have led
to market withdrawal or suspension of these agents. Post market
findings revealed that cisapride binds to the human ether-à-go-go
related gene (hERG) potassium channel, an ion channel important
for the cardiac action potential. High affinity for this channel can
give rise to prolonged QT interval and cause fatal torsades des
pointes arrhythmia [8].
* Corresponding author. School of Pharmacy, University of Oslo, P.O. Box 1068,
Blindern, N-0316 Oslo, Norway. Tel.: þ47 22855043; fax: þ47 22857975.
E-mail address: jo.klaveness@farmasi.uio.no (J. Klaveness).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.03.060

630
B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
Increased focus on hERG has revealed strategies to avoid drug
binding to this ion channel, and the incorporation of negatively-
charged functional groups seems to be a feasible way to overcome
hERG potassium channel-related side effects [9]. This strategy can
also be used to avoid distribution of peripheral-acting drugs to the
central nervous system, and it has been successfully introduced to
second-generation antihistamine H1-receptor antagonists. All
second-generation anti-allergic drugs contain a carboxylic acid
group, or they are metabolized to active metabolites with a carbox-
ylic acid group [10]. Terfenadine, a H1-receptorantagonistwith a tert-
butyl group, is extensively metabolized to the corresponding
dimethyl acetic acid metabolite fexofenadine as outlined in Fig. 1.
Fexofenadine, in contrast to terfenadine, does not bind to the hERG
channel and has limited CNS distribution because of the negatively
charged carboxylate group. However, reports of prolonged QT in-
tervals and arrhythmias due to incomplete metabolic conversion of
terfenadine to fexofenadine have led to market withdrawal of ter-
fenadine. Only the carboxylic acid metabolite fexofenadine is used
today in anti-allergic treatment. It has also been shown that the
carboxylic acid group is important for antihistamine drugs to act as
substrates for P-glycoproteins, and second generation antihista-
mines are actively transported out of the central nervous system [11].
We have previously reported the synthesis and pharmacological
effects of some novel hydrophilic 5-HT4 receptor antagonists [12].
Since 5-HT4 agonists might have beneficial effects in the central
nervous system, reduced distribution of 5-HT4 antagonists to the
CNS may be an advantage. Reducing the risk of cardiotoxic side
effects like prolonged QT interval may be achieved by increasing
the hydrophilicity of new therapeutic 5-HT4 ligands, for example by
incorporating a carboxylic acid moiety [9]. We have therefore
prepared a new hydrophilic 5-HT4 receptor antagonist 9 that has
structural similarities with the second-generation H1-antagonist
fexofenadine as shown in Fig. 2. The dimethyl acetic acid antagonist
9 shows promising 5-HT4 receptor binding and low affinity for the
hERG potassium channel. The log DOct7.4 value also indicates low
distribution to the central nervous system, a clinical benefit since
reduced CNS side effects may be expected.
9
Fexofenadine
10
11
GR113808
SB207266
Fig. 2. Structures of fexofenadine, GR113808, SB207266 and the synthesized de-
rivatives 9e11.
The synthesis of benzyl-protected piperidine derivative 3 and
the corresponding de-protected amine 4 is shown in Scheme 1.
The synthesis of indole ester 1 and piperidine methanol derivative
2 has been described elsewhere [13,14]. The benzyl protected
intermediate 3 was prepared by adding n-BuLi to a cooled solution
of piperidine methanol derivative 2 in THF, followed by indole ester
1 and stirring to room temperature. Crystallization from EtOAc
gave intermediate 3 in 89% yield. Hydrogenation of benzyl inter-
mediate 3 using Pd/C at 5 bar in a mixture of glacial acetic acid and
methanol at room temperature gave the de-protected amine 4 in
81% yield.
The synthesis of the bromide intermediate 7 is shown in Scheme
2. Alkylation of 2-(4-methylphenyl)propanoic acid with MeI in DMF
in the presence of NaHCO3 at room temperature gave the methyl
ester 5 in 95% yield. Addition of 5 to a cooled suspension of NaH in
n-hexane, followed by MeI and stirring at room temperature gave
intermediate 6 in 49% yield. Addition of N-bromosuccinimide to a
solution of 6 in dichloromethane followed by benzoyl peroxide and
heating the mixture to reflux gave the bromo derivative 7 in 61%
yield after filtration of the reaction mixture through a pad of silica.
The synthesis of carboxylic acid 9 and the corresponding prodrug
ester 10 is shown in Scheme 3. Addition of bromo derivative 7 to a
solution of amine 4 in acetone in the presence of K2CO3 and heating
the mixture to reflux gave the methyl ester 8 in 45% yield after pu-
rification with column chromatography. Hydrolysis of methyl ester 8
with sodium hydroxide in a mixture of aqueous MeOH under reflux
gave carboxylic acid 9 in 45% yield. Alkylation of carboxylic acid 9
with 4-chloromethyl-5-methyl-1,3-dioxol-2-one in DMA in the
However, the hydrophilic carboxylic acid 9 showed limited oral
bioavailability in a rat model. To increase oral bioavailability of the
new antagonist, two different prodrug approaches were applied.
First the corresponding tert-butyl prodrug 11 was synthesized and
evaluated with respect to oral bioavailability. The tert-butyl deriv-
ative 11 did not improve bioavailability, and plasma analysis
revealed incomplete metabolism to carboxylic acid 9. Therefore, the
corresponding medoxomil ester prodrug 10 was prepared. The
medoxomil ester 10 showed both complete conversion to carboxylic
acid 9 and satisfactory oral bioavailability and pharmacokinetics.
2. Chemistry
The hydrophilic carboxylic acid 9, a derivative of the 5-HT4
antagonist SB207266, and the corresponding prodrugs 10 and 11
are shown in Fig. 2.
CYP450
3A4
Terfenadine - prodrug
Fexofenadine - active metabolite
Fig. 1. First-pass metabolism of the second generation H1 receptor antagonist terfenadine to the active metabolite fexofenadine.

B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
631
b
a
1
2
3
4
Scheme 1. Reagents and conditions: a) n-BuLi, THF, 0 ꢂC ➔ room temperature; b) H2, PdeC 20%, CH3CO2H/MeOH, room temperature.
presence of K2CO3 underreflux gave the prodrug ester 10 in 50% yield
after purification with column chromatography.
criteria, and the binding affinity and adenylyl cyclase data were
measured to pKi 9.9 ꢀ 0.2 and pKb 8.6 ꢀ 0.1, respectively.
The synthesis of the tert-butyl derivative 11 is shown in Scheme
4. Alkylation of piperidine amine 4 with 4-(tert-butyl)-1-benzyl
bromide in acetone in the presence of K2CO3 under reflux gave
the tert-butyl derivative 11 in 59% yield after purification with
column chromatography.
New drug candidates are commonly tested across a panel of
side-effect targets. One of these is the affinity for the hERG potas-
sium channel. High affinity for this ion channel may indicate a risk
of prolonged QT intervals and serious cardiac arrhythmias [17].
Several in vitro tests are available to screen out possible hERG
channel blockers, but the whole cell patch-clamp technique is
regarded the most reliable method to measure interaction with the
potassium channel [18]. Compound 9 was tested by the patch-
clamp technique and the peak hERG tail amplitude was reduced
3. Pharmacology
Binding affinities and 5-HT4 antagonist properties were deter-
mined for GR113808, SB207266 and compound 9 and 11 by
competitive binding and concentration-dependent inhibition of 5-
HT-stimulated adenylyl cyclase (AC) activity in membranes from
HEK293 cells stably expressing the human 5-HT4(b) receptor. pKi-
values and pKb-values from competition of [3H] GR113808 binding
and from antagonism of 5-HT4-stimulated adenylyl cyclase activity,
respectively, for GR113808, SB207266 and compounds 9 and 11 are
summarized in Table 1.
Reduction of hERG tail current of compound 9 was quantified
with the patch-clamp technique in HEK293 cells stably transfected
with hERG ion channel cDNA. Peak hERG tail current amplitude was
measured prior to and following exposure to compound 9 at
different concentrations. Terfenadine, a known inhibitor of the IKr
current, was used as a positive control.
Oral bioavailability and pharmacokinetic properties of com-
pounds 9, 10 and 11 were determined in male Sprague-Dawley rats.
The concentration of compound 9 in plasma was measured by
liquid chromatography-mass spectrometry (LC-MS) after oral
dosing of 9, 10 and 11. To calculate the oral bioavailability, com-
pound 9 was also administered by intravenous (i.v.) dosing. The
pharmacokinetic calculations for tentative bioavailability (F, %),
maximum plasma concentration (Cmax, ng/ml), clearance (CL, ml/
min/kg), half-life (T1/2, min) and time to reach maximum plasma
concentration (Tmax, min) are shown in Table 2.
by 55.8 ꢀ 5.9% prior to and following exposure of a 100
mM solution.
M (n ¼ 3) after
The corresponding IC50 value was estimated to 69.0
m
fitting the data points with a sigmoidal function. A 50 nM solution
of terfenadine, a known inhibitor of the IKr current, inhibited
69.4 ꢀ 3.1% in the same assay. Based on binding and AC assays, the
therapeutic concentration of a 5-HT4 antagonist with intended
pharmacodynamic action in the human ventricle is expected to be
in the low-nanomolar range. Therefore, it is unlikely that com-
pound 9 will block the hERG potassium channel and provoke car-
diac arrhythmias in a clinical situation. This finding is also
consistent with previous results obtained with 5-HT4 ligands con-
taining carboxylic acid groups [12].
Screening of absorption, distribution, metabolism and excretion
(ADME) properties is important in the early discovery and lead-
optimization phases for a new drug. Orally active drugs are absor-
bed from the intestine, transported by the portal vein to the liver
before reaching the systemic circulation. After the drug enters the
systemic circulation it is distributed from the blood and into
different organs and tissues. Several variables like molecular
weight, active transport mechanisms and plasma protein binding
are important for the ADME properties of the drugs. Lipophilicity of
the drug compound is also regarded as an important factor. Several
methods are used to determine lipophilicity, but the distribution
coefficient is the most commonly used. Assuming passive diffusion,
the logarithmic distribution coefficient is often used to obtain a
rough estimate of a drugs distribution and permeation properties.
Hydrophilic drugs with negative values of log DOct7.4 are assumed to
stay in hydrophilic compartments, have less systemic toxicity and
are excreted from the body by renal clearance. More hydrophobic
drugs with positive values are more likely to be metabolized by first
pass metabolism in the liver, penetrate membranes and accumulate
in organs and tissues such as the central nervous system [19e23].
The logarithmic partition coefficient between 1-octanol and
phosphate buffer pH 7.4 was determined using the shake-flask
method to ꢁ1.2 ꢀ 0.1 for the carboxylic acid 9. The relatively
small log DOct7.4 value is caused by deprotonation of the carboxylic
acid to the more water soluble carboxylate anion. The lead com-
pound SB207266 and prodrugs 10 and 11 are significantly less
water soluble with log DOct7.4 values ranging from 2.2 ꢀ 0.1 to
2.4 ꢀ 0.1. Even though more extensive animal studies are needed to
determine the distribution and permeation patterns of these
compounds, the log DOct7.4 value indicates that the carboxylic acid 9
will tend to have reduced CNS distribution compared to SB207266.
This may be an advantage since the intended therapeutic use for a
new hydrophilic 5-HT4 antagonist will be in the cardiovascular
The in vitro stability and metabolic profile of compound 9 was
determined in human liver microsomes. Metabolic degradation of 9
and identification of metabolites was analyzed by LC-MS.
4. Results and discussion
The 5-HT4 receptor binding site is extensively described in
literature [15]. The pharmacophore antagonist model features an
aromatic ring, a coplanar carbonyl group and an alkaline nitrogen
substituted with a bulky group [16]. Carboxylic acid 9 fulfills these
b
a
c
5
6
7
Scheme 2. Reagents and conditions: a) MeI, NaHCO3, DMF, room temperature; b) MeI,
NaH, n-hexane, room temperature; c) NBS, BPO, CH2Cl2, reflux.

632
B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
b
a
8
4
7
c
9
10
Scheme 3. Reagents and conditions: a) K2CO3, acetone, reflux; b) NaOH, MeOH/H2O, reflux; c) 4-chloromethyl-5-methyl-1,3-dioxol-2-one, K2CO3, DMA, reflux.
system of heart failure patients [5,6]. Hydrophilic 5-HT4 antagonists
may also have therapeutic benefits in other peripheral organs and
tissues like the intestine and urinary system. Further, it has been
shown that stimulation of 5-HT4 receptors seems to play a vital role
in cognitive function through the release of acetylcholine and
amyloid precursor protein (APP) peptides [24,25]. Reduced distri-
bution into the CNS may therefore be important to avoid unwanted
side effects of a new 5-HT4 receptor antagonist.
The pharmacokinetic parameters for compounds 9, 10 and 11
are summarized in Table 2, and the corresponding plasma con-
centrations versus time curves are shown in Fig. 3. To evaluate the
potential for new compounds to be orally active, a threshold area
under the curve (AUC) value of 500 h*ng/ml after a single 10 mg/kg
oral dose to rat has been proposed [26]. The obtained total oral AUC
for compound 9 was 285 ꢀ 10 h*ng/ml, and the calculated tentative
oral bioavailability 3%.
A strategy to increase oral bioavailability for H1-receptor an-
tagonists is shown in Fig. 1. The tert-butyl prodrug terfenadine is
oxidized in the liver to the dimethyl acetic acid fexofenadine.
Therefore, to increase oral bioavailability of dimethyl acetic acid 9,
we decided to use the same strategy and the tert-butyl prodrug 11
was prepared. However, the obtained total AUC for carboxylic acid 9
after administration of prodrug 11 was only 127 ꢀ 16 h*ng/ml,
corresponding to a bioavailability of 1%.
Unlike the second generation H1-receptor antagonist terfena-
dine, where the tert-butyl group is completely metabolized to the
active dimethyl acetic acid metabolite fexofenadine, LC-MS analysis
revealed considerably amounts of unmetabolized 11 in plasma. The
AUC for 11 was 533 ꢀ 65 h*ng/ml, corresponding to a bioavailability
of 6%. Despite incomplete metabolic oxidation to carboxylic acid 9,
the low AUC value may indicate extensive liver degradation of 11. It
was concluded that the tert-butyl prodrug strategy, as applied for
fexofenadine, was not applicable for the dimethyl acetic acid 9.
The maximum plasma concentration of compound 9 was
reached at 12 ꢀ 3 and 80 ꢀ 20 min after administration of com-
pound 9 and 11, respectively. The difference in Tmax is not unex-
pected since compound 11 needs to be metabolized to the acid 9
before reaching the circulatory system. A contributing factor can
also be the possible lower solubility of compound 11 in the rat in-
testine. The calculated elimination half-life for compound 9 after
oral administration of 9 and 11 was 364 ꢀ 113 and 1190 ꢀ 339 min,
respectively. The difference in elimination half-life also indicates
that the absorption and conversion phase has an impact on the
elimination of the carboxylic acid 9. The plasma clearance for
compound 9 is 17 ꢀ 1 ml/min/kg and is in the intermediate range
when compared to typical rat liver blood flow of 55 ml/min [27].
The theoretical maximal bioavailability (Fmax) for compound 9
would therefore be:
Fmax ¼ ð1 ꢁ Clearance=Liver blood flowÞ ꢃ 100 %
¼ ð1 ꢁ 17:2 ml=min=55 ml=minÞ ꢃ 100 % ¼ 68 %:
The reasons for lower bioavailability can be degradation in the
gastrointestinal tract, poor permeability or extensive first-pass
metabolism. The relatively low Tmax eliminates that poor perme-
ability would be the only reason for low bioavailability. More likely
the observed pharmacokinetics could be explained by extensive
first-pass metabolism. This has also been seen in previous studies
with 5-HT4 ligands with aromatic ester functionalities [28].
To determine if the limited oral bioavailability was due to
extensive liver degradation, the in vitro stability of compound 9 in
human liver microsomes was evaluated. The concentration of car-
boxylic acid 9 in the in vitro assay was followed by LC-MS, indi-
cating negligible metabolic degradation.
The microsome assay was also screened for potential degrada-
tion products, and two metabolites were detected. The abundance
of these metabolites was low, less than 0.1% in the 60 min incu-
bation sample. LC-time-of-flight (TOF)-MS analysis identified the
degradation products as mono- and di-hydroxylated metabolites,
where the hydroxyl groups were situated in the indole-oxazino
ring system. Similar hydroxylated metabolites have previously
been reported for the lead compound SB207266 [14].
Since the in vitro disappearance is slow and the predicted liver
metabolic clearance negligible, it can be assumed that the limited
bioavailability in rat is not solely caused by extensive first-pass
metabolism. The observed difference between in vitro and in vivo
a
4
11
Scheme 4. Reagents and conditions: a) K2CO3, acetone, reflux.

B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
633
Table 1
Logarithmic distribution coefficient (log DOct7.4) from an octanol/buffer distribution,
pKi-values from competition of [3H] GR113808 binding, pKb-values from antagonism
of 5-HT-stimulated adenylyl cyclase activity of GR113808, SB207266 and compounds
9e11.
Compound log DOct7.4 ꢀ SEMa pKi ꢀ SEMb pKb ꢀ SEMc Molecular formula
GR113808 ND
10.4 ꢀ 0.2 8.7 ꢀ 0.1
10.0 ꢀ 0.1 9.1 ꢀ 0.1
9.9 ꢀ 0.2 8.6 ꢀ 0.1
C19H27N3O4S
C22H31N2O2
C29H34N2O5
SB207266
9
10
11
2.2 ꢀ 0.1
ꢁ1.2 ꢀ 0.1
2.2 ꢀ 0.1
2.4 ꢀ 0.1
ND
ND
ND
ND
C34H38N2O8 ꢃ HCl
C29H36N2O3 ꢃ HCl
a
n ¼ 3.
b
n ¼ 3e8.
c
n ¼ 4e7.
clearance is probably due to species differences in metabolism,
active excretion to urine/bile or some non-liver degradation
pathway. Despite the low Tmax value of dimethyl acetic acid 9,
reduced absorption from the gastrointestinal tract may be a plau-
sible explanation of limited oral bioavailability.
Preparation of ester prodrugs to increase passive diffusion
through the intestinal wall is a well-known strategy [29]. The
medoxomil ester prodrug is used to increase oral bioavailability of
olmesartan, an angiontensin II receptor antagonist used in treat-
ment of high blood pressure [30]. The olmesartan medoxomil ester
prodrug has an oral bioavailability of 25% in humans. The medox-
omil ester is completely metabolized to olmesartan carboxylic acid
in the intestinal wall and liver, and no other component than the
active olmesartan metabolite has been detected in plasma after oral
administration [31,32]. To increase oral bioavailability of 9 and to
minimize systemic concentrations of prodrug, the medoxomil ester
prodrug 10 was synthesized and evaluated in the same animal
model.
Fig. 3. Mean plasma concentrationetime profiles of compound 9 after oral adminis-
tration of compound 9 (-; n ¼ 3), 10 (:; n ¼ 2) and 11 (C; n ¼ 3) at a dose of 10 mg/
kg in rats.
mediated adenylyl cyclase activity. The hydrophilic derivative has
minor affinity for the hERG ion channel and with a partition coef-
ficient of ꢁ1.2, reduced CNS accumulations may be expected. The
limited oral bioavailability can be overcome by administration of
the medoxomil ester prodrug 10. Further studies with the hydro-
philic 5-HT4 antagonist 9 as a new therapeutic opportunity in
treatment of heart failure are planned in the future.
The pharmacokinetic parameters for compound 10 are sum-
marized in Table 2, and the corresponding plasma concentration
versus time curve for compound 9 is shown in Fig. 3. The total oral
AUC for carboxylic acid 9 was 1206 h*ng/ml after administration of
prodrug 10, well above the threshold value of 500 h*ng/ml. The
tentative oral bioavailability increased from 3% for the dimethyl
acetic acid 9 to 16% for the corresponding medoxomil ester prodrug
10. The stability of 9 in liver microsomes suggests that the oral
bioavailability may be higher in other animal species or humans.
The maximum peak plasma concentration and elimination half-life
was found to be 30 and 72 min respectively. The medoxomil pro-
drug 10 was not detected in any of the harvested plasma samples.
6. Experimental section
6.1. Chemistry
1H NMR spectra were recorded on a Bruker Spectrospin Avance
spectrometer at 300 MHz for 1H and 75 MHz for 13C. Chemical shifts
are reported in parts per million relative to internal tetrame-
thylsilane, residual CHCl3
(dH ¼ 7.26 ppm, dC ¼ 77.0 ppm) or
CHD2SOCD3
(dH ¼ 2.50 ppm, dC ¼ 39.43 ppm). Coupling constants
(J) are reported in hertz (Hz). The following abbreviations are used
to describe peak patterns when appropriate: s (singlet), d (doublet),
t (triplet), q (quartet), m (multiplet) and br (broad). Analytical thin-
layer chromatography (TLC) was run on Merck silica gel plates
(Kieselgel 60 F-254) with UV-light or iodine detection. For flash
chromatography, Fluka silica gel type 60 (size 200e400 mesh) was
used. All solvents and reagents were of analytical or reagent grade
and were obtained from commercial sources.
5. Conclusion
We have synthesized a new hydrophilic 5-HT4 receptor antag-
onist 9 with a benzyl dimethyl acetic acid group. The new antag-
onist shows excellent receptor binding and antagonism of 5-HT4-
(1-Benzyl-4-piperidinyl)methanol 1 and methyl 3,4-dihydro-
2H-[1,3]oxazino[3,2-a]-indole-10-carboxylate 2 were synthesized
according to literature procedures [13,14].
Table 2
Pharmacokinetic (PK) data for carboxylic acid 9 after administration of compound
9e11 in rats.
6.1.1. Preparation of (1-benzyl-4-piperidinyl)methyl 3,4-dihydro-
2H-[1,3]oxazino[3,2-a]indole-10-carboxylate (3)
Compound AUC
Cmax
Tmax
T1/2
CL
%F
(hr*ng/ml) (ng/ml)
(min)
(min)
(ml/min/kg)
1.6 M n-BuLi in THF (25.3 ml, 40.5 mmol) was added dropwise to
a solution of piperidine methanol 2 (8.35 g, 40.5 mmol) in THF
(50 ml) at 0 ꢂC under argon atmosphere and stirred for 5 min.
Methyl ester 1 (6.26 g, 27.1 mmol) was then added and the reaction
mixture stirred at room temperature for 12 h. The reaction mixture
was carefully hydrolyzed with water (30 ml) and extracted with
CH2Cl2 (100 ml). The organic layer was washed with water
9a
285 ꢀ 10 130 ꢀ 40 12 ꢀ 3
364 ꢀ 113 17 ꢀ 1
72 NA
1190 ꢀ 339 18 ꢀ 3
3
16
1
10b
11c
11d
1206
817
30
127 ꢀ 16 13 ꢀ 4
80 ꢀ 20
533 ꢀ 65 33 ꢀ 5
340 ꢀ 140 1610 ꢀ 483 20 ꢀ 5
6
a
PK of 9 after p.o. administration of 9 (n ¼ 3).
PK of 9 after p.o. administration of 10 (n ¼ 2).
PK of 9 after p.o. administration of 11 (n ¼ 3).
PK of 11 after p.o. administration of 11 (n ¼ 3).
b
c
d
(3
ꢃ
30 ml), dried over anhydrous Na2SO4, filtered and

634
B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
concentrated in vacuo to leave an oil. The residue was treated with
(0.62 g, 4.5 mmol) in acetone (10 ml) and heated at reflux for 24 h.
The mixture was cooled to room temperature, filtered and the
filtrate evaporated in vacuo. The residue was redissolved in CH2Cl2
(20 ml) and washed with H2O (3 ꢃ 5 ml). The organic layer was
dried over Na2SO4, filtered and evaporated in vacuo. The residue
was separated with flash chromatography (CH2Cl2/MeOH 9:1) to
leave the title compound 8 as a white solid (0.34 g, 45%). 1H NMR
EtOAc and the precipitate filtered off to leave the title compound 3
as a white solid (9.70 g, 89%). 1H NMR (CDCl3):
d 7.96e7.93 (m, 1H),
7.31e7.10 (m, 8H), 4.50 (t, J ¼ 5.2 Hz, 2H), 4.17 (d, J ¼ 6.2 Hz, 2H),
4.06 (t, J ¼ 6.2 Hz, 2H), 3.49 (s, 2H), 2.93e2.87 (m, 2H), 2.32e2.27
(m, 2H), 2.05e1.78 (m, 5 H), 1.45e1.39 (m, 2H). MS (ES): 405.2
[M þ H]þ.
(CDCl3):
d 7.97e7.92 (m, 1H), 7.25‒7.16 (m, 5H), 7.14‒7.10 (m, 2H),
6.1.2. Preparation of 4-piperidinylmethyl 3,4-dihydro-2H-[1,3]
oxazino[3,2-a]indole-10-carboxylate (4)
4.51 (t, J ¼ 5.2 Hz, 2H), 4.17 (d, J ¼ 6.2 Hz, 2H), 4.08 (t, J ¼ 6.2 Hz, 2H),
3.63 (s, 3H), 3.47 (s, 2H), 2.92‒2.86 (m, 2H), 2.37‒2.29 (m, 2H), 2.02‒
1.92 (m, 2H), 1.91‒1.66 (m, 3H). 1,55 (s, 6H), 1.43‒1.38 (m, 2H). MS
(ES): 505.2 [M þ H]þ.
A solution of benzyl amine 3 (9.70 g, 24.0 mmol) in a mixture of
glacial acetic acid (15 ml) and MeOH (80 ml) was hydrogenated
over 20% Pd/C (1.0 g) and 5 bar at room temperature for 48 h. The
reaction mixture was filtered and the filtrate diluted with H20
(30 ml). The filtrate was made alkaline with K2CO3 to pH 11 and the
aqueous mixture extracted with CH2Cl2 (3 ꢃ 50 ml). The organic
layers were combined and dried over anhydrous Na2SO4, filtered
and evaporated in vacuo to leave the title compound 4 as a white
6.1.7. Preparation of 2-[4-[[4-(3,4-dihydro-2H-[1,3]oxazino[3,2-a]
indole-10-carbonyloxymethyl)-1-piperidinyl]methyl]phenyl]-2-
methylpropanoic acid (9)
Methyl ester 8 (0.34 g, 0.67 mmol) was added to a mixture of
2 M aqueous NaOH (1 ml) and MeOH (4 ml) and stirred at reflux for
12 h. The mixture was cooled to room temperature and evaporated
in vacuo. The residue was redissolved in H2O (5 ml) and the solu-
tion acidified to pH 3 with 2 M aqueous HCl. The free carboxylic acid
precipitate was filtered off and the residue recrystallized from
acetone to leave the title compound 9 as a white solid (0.15 g, 46%);
solid (6.12 g, 81%). 1H NMR (DMSO-d6):
d 7.83e7.79 (m, 1H), 7.28‒
7.26 (m, 1H), 7.14‒7.07 (m, 2H), 4.51 (t, J ¼ 4.7 Hz, 2H), 4.17‒4.04 (m,
4H), 3.26‒3.20 (m, 2H), 2.86‒2.80 (m, 2H), 2.27‒2.22 (m, 2H), 1.95‒
1.81 (m, 3H), 1.60‒1.50 (m, 2H). MS (ES): 315.2 [M þ H]þ.
6.1.3. Preparation of methyl 2-(4-methylphenyl)propanoate (5)
MeI (8.64 g, 76.1 mmol) was added dropwise to a mixture of 2-
(4-methylphenyl)propanoic acid (5.00 g, 30.5 mmol) and NaHCO3
(6.39 g, 76.1 mmol) in DMF (50 ml) and stirred at room temperature
for 48 h. The reaction mixture was poured into ice water (100 ml).
The aqueous solution was acidified to pH 3 with concentrated HCl
and the aqueous mixture extracted with Et2O (3 ꢃ 50 ml). The
combined organic layers were dried over anhydrous Na2SO4,
filtered and evaporated in vacuo to leave the title compound 5 as a
mp. 235e238 ꢂC. 1H NMR (DMSO-d6):
d 12.10 (br s, 1 H), 7.81 (d,
J ¼ 5.9 Hz, 1H), 7.31‒7.16 (m, 5H), 7.13‒7.06 (m, 2H), 4.48 (t,
J ¼ 4.7 Hz, 2H), 4.10 (t, J ¼ 6.0 Hz, 2H), 4.02 (d, J ¼ 5.7 Hz, 2H), 3.38 (s,
2H), 2.85‒2.80 (m, 2H), 2.26‒2.21 (m, 2H), 2.01‒1.91 (m, 2H), 1.72‒
1.67 (m, 3H), 1.44 (s, 6H), 1.32‒1.22 (m, 2H). 13C NMR (DMSO-d6):
d
164.0, 154.8, 152.7, 132.1, 131.7, 127.7, 126.3, 125.5, 122.8, 121.2,
120.1, 109.7, 85.6, 80.0, 79.9, 67.4, 66.4, 59.6, 51.7, 35.2, 34.1, 31.8,
26.5, 21.1. HRMS (TOF MS ESþ) for C29H34N2O5 [M þ H]þ: calcd.
491.2468 found 491.2460.
yellow oil (5.10 g, 95%). 1H NMR (CDCl3):
d 7.19‒7.08 (m, 4H), 3.72‒
3.65 (m, 1H), 3.63 (s, 3H), 2.30 (s, 3H), 1.46 (d, J ¼ 7.2 Hz, 3H). MS
6.1.8. Preparation of [1-[[4-[1,1-dimethyl-2-[(5-methyl-2-oxo-1,3-
dioxol-4-yl)methoxy]-2-oxo-ethyl]phenyl]methyl]-4-piperidinyl]
methyl 3,4-dihydro-2H-[1,3]oxazino[3,2-a]indole-10-carboxylate
hydrochloride (10)
(EI): 179.1 [M þ H]þ.
6.1.4. Preparation of methyl 2-methyl-2-(4-methylphenyl)
propanoate (6)
4-Chloromethyl-5-methyl-1,3-dioxol-2-one (78 mg, 0.53 mmol)
was added to a suspension of carboxylic acid 9 (0.20 g, 0.40 mmol)
and K2CO3 (0.16 g, 1.2 mmol) in dimethylacetamide (1.0 ml). The
mixture was stirred at room temperature for 12 h, evaporated in
vacuo and the residue separated by flash chromatography
(CH2Cl2:MeOH 9:1) to leave the title compound 10 as a white solid
(0.12 g, 50%). The corresponding hydrochloride salt was prepared
by dissolving the title compound 10 in CH2Cl2 and adding ethereal
HCl. After stirring for 1 h, the mixture was evaporated in vacuo and
the residue recrystallized from acetone to leave the hydrochloride
salt as a white crystalline solid; mp. 224e227 ꢂC. 1H NMR (DMSO-
A solution of methyl ester 5 (5.10 g, 28.6 mmol) inTHF (30 ml) was
added dropwise to a suspension of NaH (5.90 g, 0.148 mol) in n-
hexane (30 ml) under argon atmosphere and stirred at room tem-
perature for 30 min. MeI (14.6 g, 0.102 mol) was added to the reaction
mixture and stirred at room temperature for 24 h. The reaction
mixture was added to H2O (50 ml) and extracted with Et2O
(3 ꢃ 50 ml). The organic layers were combined and dried over
anhydrous Na2SO4, filtered and evaporated in vacuo to leave the title
compound 6 as a yellow oil (2.72 g, 49%).1H NMR (CDCl3):
d 7.23‒7.09
(m, 4H), 3.63 (s, 3H), 2.31 (s, 3H),1.56 (s, 6H). MS (EI): 193.1 [M þ H]þ.
d6):
d
10.30 (br s, 1 H), 7.80 (d, J ¼ 6.7 Hz, 1H), 7.58‒7.52 (m, 2H),
6.1.5. Preparation of methyl 2-methyl-2-(4-bromomethylphenyl)
propanoate (7)
7.37‒7.28 (m, 2H), 7.15‒7.06 (m, 2H), 4.97 (s, 2H), 4.50 (t, J ¼ 4.7 Hz,
2H), 4.22‒4.21 (m, 2H), 4.11 (t, J ¼ 5.9 Hz, 2H), 4.04 (d, J ¼ 5.8 Hz,
2H), 3.31 (s, 2H), 2.94‒2.91 (m, 2H), 2.26‒2.22 (m, 2H), 2.11 (s, 3H),
1.93‒1.89 (m, 3H), 1.63‒1.55 (m, 2H), 1.51 (s, 6H). 13C NMR (DMSO-
N-bromosuccinimide (3.00 g, 16.9 mmol) and benzoyl peroxide
(catalytic amount) was added to a solution of intermediate 6
(2.72 g, 14.1 mmol) in CH2Cl2 (30 ml) and heated at reflux for 12 h.
The reaction mixture was cooled to room temperature and filtered
through a pad of silica gel. The filtrate was evaporated in vacuo to
leave the title compound 7 as a yellow oil (2.32 g, 60%). 1H NMR
d6): d 163.8, 154.6, 152.5, 149.9, 137.7, 134.4, 132.0, 131.7, 127.6, 126.2,
125.4, 122.8, 121.1, 120.1, 109.5, 85.4, 80.0, 79.7, 67.2, 66.2, 59.5, 51.6,
35.1, 34.0, 33.3, 31.6, 26.3, 21.0, 9.7. HRMS (TOF MS ESþ) for
C
34H38N2O8 [M þ H]þ: calcd. 603.2706 found 603.2699.
(CDCl3):
d 7.33‒7.23 (m, 4H), 4.46 (s, 2H), 3.63 (s, 3H), 1.55 (s, 6H).
MS (EI): 271.0 [M þ H]þ.
6.1.9. Preparation of [1-[(4-tert-butylphenyl)methyl]-4-piperidinyl]
methyl 3,4-dihydro-2H-[1,3]oxazino[3,2-a]indole-10-carboxylate
hydrochloride (11)
4-(tert-Butyl)-1-benzyl bromide (0.25 g, 1.0 mmol) was added to
a stirred suspension of piperidine amine 4 (0.31 g, 1.0 mmol) and
K2CO3 (0.41 g, 3.0 mmol) in acetone (10 ml) and heated to reflux for
24 h. The mixture was cooled to room temperature, filtered and the
6.1.6. Preparation of [1-[[4-(2-methoxy-1,1-dimethyl-2-oxo-ethyl)
phenyl]methyl]-4-piperidinyl]methyl 3,4-dihydro-2H-[1,3]oxazino
[3,2-a]indole-10-carboxylate (8)
Bromomethyl ester 7 (0.40 g, 1.5 mmol) was added to a stirred
suspension of piperidine amine 4 (0.47 g, 1.5 mmol) and K2CO3

B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
635
filtrate evaporated in vacuo. The residue was redissolved in CH2Cl2
(20 ml) and washed with H2O (3 ꢃ 5 ml). The organic layer was
dried over Na2SO4, filtered and evaporated in vacuo. The residue
was separated with flash chromatography (CH2Cl2/MeOH 9:1) to
leave the title compound 11 as a white solid (0.27 g, 59%). The
corresponding hydrochloride salt was prepared by dissolving the
title compound 11 in CH2Cl2 and adding ethereal HCl. After stirring
for 1 h, the mixture was evaporated in vacuo and the residue
recrystallized from acetone to leave the hydrochloride salt as a
white crystalline solid; mp. 216e220 ꢂC. 1H NMR (DMSO-d6):
The whole-cell configuration was achieved first by forming a
giga-ohm seal between the cell membrane and the patch pipette
using gentle suction and then by rupturing the membrane patch
under the electrode tip with a stronger suction pulse. The cell
membrane potential was then clamped at ꢁ80 mV. Initial leak
current was <100 pA and was subtracted using the amplifier and
then continuously monitored at the holding potential (ꢁ80 mV)
with an oscilloscope. Whole-cell patch-clamp recordings were
performed using an Axopatch 200B amplifier (Molecular Devices,
CA, USA) linked to a computer equipped with pCLAMP software
(version 10.1). Data were filtered at 2 kHz and sampled using a
Digidata analog-to-digital board.
The standard voltage protocol was run at 15 second intervals as
follows: A step from e80 to þ20 mV for 2 seconds activated hERG
channels generating an outward current. A step from þ20 to e50
mV for 2 seconds generated the tail current on which the effect of
the test compound was determined. A step back to e80 mV
returned the cell to the holding potential.
d
10.60 (s, 1H), 7.82-7.79 (m, 1H), 7.56‒7.42 (m, 4H), 7.30 (d,
J ¼ 6.7 Hz, 1H), 7.17‒7.08 (m, 2H), 4.50 (t, J ¼ 4.7 Hz, 2H), 4.20‒4.18
(m, 2H), 4.10 (t, J ¼ 5.9 Hz, 2H), 4.03 (d, J ¼ 5.8 Hz, 2H), 3.33‒3.30 (m,
2H), 2.94‒2.89 (m, 2H), 2.27‒2.23 (m, 2H), 1.93‒1.87 (m, 3H), 1.74‒
1.63 (m, 2H), 1.27 (s, 9H). 13C NMR (DMSO-d6):
d 154.8, 152.7,
132.1131.7, 127.7, 126.3, 125.5, 122.8, 121.2, 120.1, 109.7, 85.6, 67.4,
66.4, 59.6, 51.8, 35.2, 34.2, 31.9, 28.0, 26.5, 21.1, 21.0. HRMS (TOF
MS ESþ) for C29H36N2O3 [M þ H]þ: calcd. 461.2803 found 461.2809.
6.4.1. Treatment groups
6.2. Log DOct7.4 measurements
Groups of cells were treated with vehicle solution, a single
concentration of terfenadine and four concentrations of compound
9. After a few minutes of equilibration, the voltage protocol was run
a minimum of ten times to determine the tail current amplitude
baseline before application of the vehicle, terfenadine or compound
Compounds 9, 10 and 11 (approximately 100 mg) were added to
phosphate buffer pH 7.4 (50 ml) and the resulting solution filtered
through a 0.43 mm filter. The phosphate buffer solutions (20 ml)
were transferred to a separation funnel and added 1-octanol
(20 ml), agitated for 24 h and set to reach equilibrium for 72 h at
room temperature. The concentration of compounds 9, 10 and 11 in
samples of phosphate buffer and 1-octanol (n ¼ 3) was quantified
with LC-UV (Agilent Technologies 1100 series, equipped with
degasser, autosampler, column oven and DAD detector). The loga-
rithmic distribution coefficient was obtained using the following
equation:
9. Ascending concentrations of compound
9 were applied
sequentially and without washout between successive concentra-
tions. Each concentration was applied until current inhibition had
reached
a steady state level. In order to construct the
concentration-response curve for compound 9, the process was
repeated on several cells until at least 3 data points per concen-
tration were obtained.
log DOct7;4 ¼ log COct=CpH7:4
6.4.2. Data analysis
The data from the hERG current analysis were performed using
the Clampfit module of pCLAMP software (version 10.2) and
Microsoft Excel 2003. The peak hERG tail current amplitude was
measured relative to the holding current at ꢁ80 mV and extracted
into a spreadsheet where the tail current amplitude was calculated
from an average of 5e10 voltage pulses. In the vehicle control
group, the amplitude of the tail current over a period of approxi-
mately 22e23 min was recorded. For each cell, the tail current
amplitude versus time was normalized against the first recorded
current amplitude value. The data were then averaged for the three
cells. For compound 9, the peak amplitude of the tail current after
exposure to each test concentration was measured and the per-
centage residual current calculated (% Baseline). To determine a
concentrationeresponse relationship for the effects of compound 9
on hERG tail current, the residual current calculated for each test
concentration was individually corrected for the mean effect of
vehicle and rundown observed using the corresponding (or closest
minute) time-matched vehicle values as follows:
where COct is the sum of ionized and unionized forms of compound
9, 10 and 11 in 1-octanol and CpH7.4 the corresponding concentra-
tions in phosphate buffer pH 7.4.
6.3. Receptor binding and adenylyl cyclase assay
The experimental procedures to determine the competitive
binding ([3H]GR113808, 0.3 nM) and concentration-dependent in-
hibition of 5-HT-stimulated (1 mM) AC activity in membranes from
HEK293 cells stably expressing the human 5-HT4(b) receptor have
previously been described [12].
6.4. Effect on hERG tail current
The experimental work was carried out at Huntingdon Life
Sciences (Cambridgeshire, England). Stock solutions of compound 9
and terfenadine (a known inhibitor of Ikr-current) were prepared in
100% DMSO. Test solutions were prepared in HEPES-buffered
physiological external salt solutions (HB-PSS) with a final perfu-
sion concentration of 0.1% DMSO.
% Baselineðvehicle correctedÞ
% Baselineðtest substanceÞ
Cells were cultured at 37 ꢀ 1 ꢂC in a humidified, gassed (set at 5%
CO2) incubator under standard conditions using Modified Eagle Me-
dium (MEM þ Glutamax) supplemented with 10% fetal bovine serum,
¼
ꢃ 100
mean % Baselineðtime matched vehicleÞ
The corrected data were then averaged for each test concen-
tration of compound 9, plotted against nominal concentrations and
the concentrationeresponse curve was fitted using the following
equation:
1% non-essential amino acids and 500 mg/ml geneticin. For patch-
clamp experiments, cells were replated on sterile glass coverslips at
low density to enable isolated cells to be selected for recording. Cells
were incubated for 2e3 days prior to recording. Coverslips were
transferred to a recording chamber mounted on the stage of an
inverted microscope. The cells were continuously superfused with
HB-PSS and only one cell per coverslip was used for experiments.
ðTop ꢁ BottomÞ
% IBaseline ¼ Bottom þ
1 þ 10nðctpꢁxÞ

636
B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
where x is the logarithm of the concentration of 9, n is the Hill
coefficient, Top and Bottom are respectively the top and bottom
plateau and ctp is the logarithm of concentration of 9 at the curve
turning point. The IC50 values for compound 9, which is defined as
the concentration at which hERG tail current is blocked by 50% after
vehicle correction, were calculated from the above equation after
the variables had been obtained from the fitting procedure.
bioavailability (F) was calculated by dividing the dose normalized
AUC after p.o. administration by the dose normalized AUC after i.v.
administration, i.e.
F ¼ ½ðAUCðp:o:Þ=Doseðp:o:ÞÞ=ðAUCði:vÞ=Doseði:v:ÞÞꢄ;
and reported as percentages (%). As compounds 10 and 11 are
prodrugs of 9, only the pharmacokinetic parameters for the free
acid 9 were calculated. Since the number of animals for each time
point for compound 10 was two, no statistical analysis of the results
was carried out.
6.5. Bioavailability and human liver microsomal stability studies
The bioavailability study was approved by the Animal Care and
Use Committee of the State Provincial Office of Southern Finland.
The experimental work for the bioavailability and microsomal
stability studies were carried out at Novamass Ltd (Oulu, Finland)
and Toxis Ltd Oy (Turku, Finland).
6.5.3. LC-MS for the bioavailability study
The quantitative LC-MS/MS data from the plasma samples were
acquired witha Waters Acquity UPLC (Waters Corp.,Milford,MA, USA)
with an autosampler, a column oven and a vacuum degasser; together
with a Waters Quattro Premier triple quadrupole mass spectrometer
equipped with a Z-spray electrospray ion source. The analytical col-
6.5.1. Animal model and sample preparations
Compounds 9 and 11 were dissolved in a solution of 10% ethanol
in sterile water to a final concentration of 1.0 mg/ml. Compound 10
umn used was a Waters BEH Shield RP18, (2.1 ꢃ 50 mm, 1.7
mm)
was dissolved in a 2% (w/v) aqueous solution of
b
-cyclodextrin to a
together with an on-line filter. The eluents were 0.1% acetic acid (A, pH
3.2) and methanol (B). A linear gradient elution with profile 10%e10%
e 90%e90% B in 0e1e2e2.5 min was employed, followed by column
equilibration. The flow rate was 0.5 ml/min and the column oven
temperature was 35 ꢂC. Data was acquired using a multiple reaction
monitoring (MRM) mode using positive ionization mode of electro-
spray. The first 1.0 min of the run was directed into waste by using a
divert valve, to decrease the ion source contamination by early eluting
matrix constituents. The mass spectrometer and UPLC system were
operated under MassLynx 4.1 software.
final concentration of 1.0 mg/ml. Compounds 9, 10 and 11 were
dosed into male Sprague Dawley rats at 10 mg/kg for per-oral (p.o.)
administration and for compound 9 at 1 mg/kg for i.v. adminis-
tration. Blood samples were collected in CapijectÒ EDTA-K2 tubes
from the lateral tail vein and by cardiac puncture (terminal sam-
ples) under isoflurane (4%) anesthesia at predefined time points
(for compounds 9 and 11 three animals were used; p.o./i.v.: 5 min,
15 min, 30 min, 1 h, 2 h, 4 h, 8 h and 24 h, for compound 10 two
animals were used; p.o.: 15 min, 30 min 1 h, 2 h, 4 h and 6 h and i.v.:
15 min, 30 min, 1 h, 2 h and 6 h). The exact time-points of dosing
and sampling were recorded. The plasma was separated via
centrifugation and frozen in ꢁ20 ꢂC, shipped to the Novamass
laboratory in dry ice and stored in ꢁ80 ꢂC until analysis.
6.5.4. Stability in human liver microsomal incubations
The pooled liver microsomes used contained 20 mg protein per
ml and consisted of liver samples from 30 donors of both genders
((BD Biosciences Discovery Labware, Woburn, MA, USA). Com-
pound 9 was incubated with liver microsomes in the presence of
appropriate cofactors. The basic incubation mixture (250 ml) con-
sisted of the following components: 0.5 mg of microsomal protein
The plasma samples were thawed at room temperature, acidi-
fied with 1% formic acid (10% v/v) and protein precipitation per-
formed with addition of acetonitrile (1:2 ratio of plasma to
acetonitrile). The plasma samples were centrifuged for 10 min at
16 100 ꢃ g at room temperature (Eppendorf 5415D, Eppendorf AG,
Hamburg, Germany) and the supernatants collected for analysis.
The standard samples were similarly prepared, and blank plasma
samples were spiked to analyte concentrations of 0.5, 1, 2, 5, 10, 20,
50, 100, 200, 500 (compounds 10 and 11) and 1000 ng/ml (com-
pound 9). The determination coefficients (R2) obtained were over
0.997 for all compounds. The plasma detection limit was less than
0.2 ng/ml for compounds 9 and 11 and 3 ng/ml for compound 10.
No peaks were detected after injection of blank rat plasma, sug-
gesting good specificity of detection. The limit of quantitation was
0.5 ng/ml for compounds 9 and 11 and 5 ng/ml for compound 10.
The back-calculated accuracies were 85e125% for all compounds.
The precisions were less than 11% for compounds 9 and 11 (n ¼ 3)
and the Snedecor-precision less than 7% (n ¼ 2) for compound 10.
per ml, compound 9 dissolved in DMSO, 1 mM NADPH and 1 mM
UDPGA. The final concentration of compound 9 was 2 mM. Two
parallel incubates, one with cofactors and one without, were
employed. Final amount of DMSO in the incubation was 1% (v/v).
Each reaction mixture was preincubated for 2 min at þ 37 ꢂC in a
shaking incubator block. Reaction was started by addition of
NADPH and UDPGA. After incubation periods of 0, 10, 30 and
60 min, 50 ml samples were collected (without cofactors time points
0 and 60 min only) and the reaction terminated by adding an equal
volume of ice-cold acetonitrile. Samples were subsequently cooled
in an ice bath and the tubes were stored at ꢁ18 ꢂC until analysis.
6.5.5. LC-MS for the stability study
The incubation samples were thawed at room temperature,
shaken and centrifuged for 10 min at 16,100 ꢃ g and the supernatants
pipetted to Maximum Recovery vials (Waters Corporation, Milford,
MA, USA). The instrumental set up and analytical method was similar
as for the bioavailability study. The flow was directed to MS via an
Acquity photo-diode-array (PDA) detector. LC-TOF-MS data was
recorded with a Waters LCT Premier XE time-of-flight mass spec-
trometer (WatersCorp., Milford, MA, USA)equippedwith a LockSpray
electrospray ionization source. A positive ionization mode of elec-
trospray was used with a cone voltage of 60 V. The mass range of m/z
140e800 was acquired. Leucine enkephalinwas used asthe lock mass
compound ([M þ H]þ m/z 556.2771) for accurate mass measure-
ments. First 0.4 min of the run was directed into waste by using a
divert valve, to decrease the ion source contamination byearlyeluting
matrix constituents. No quantification was performed. The
6.5.2. Pharmacokinetic calculations
The pharmacokinetic parameters were calculated by WinNonlin
Pro (Pharsight Corp., CA, USA) using standard non-compartment
methods. Data from each animal was fitted individually and the
resulting pharmacokinetic parameters were then averaged. The
elimination phase half-life (T1/2) was calculated by least-squares
regression analysis of the terminal linear part of the concentra-
tionetime curve. The area under the plasma concentrationetime
curve was determined by use of the linear trapezoidal rule up to the
last measurable concentration and thereafter byextrapolation of the
terminal elimination phase to infinity. The maximum plasma con-
centration (Cmax) and the time to Cmax (Tmax) were derived directly
from the plasma concentration data. The tentative oral

B. Brudeli et al. / European Journal of Medicinal Chemistry 64 (2013) 629e637
637
[8] F.D. Ponti, E. Poluzzi, A. Cavalli, M. Recanatini, N. Montanaro, Drug Saf. 25
(2002) 263e286.
[9] B. Zhu, Z.J. Jia, P. Zhang, T. Su, W. Huang, E. Goldman, D. Tumas, V. Kadambi,
P. Eddy, U. Sinha, R.M. Scarborough, Y. Song, Bioorg. Med. Chem. Lett. 16
(2006) 5507e5512.
[10] F. Estelle, R. Simons, New Engl. J. Med. 351 (2004) 2203e2217.
[11] C. Chen, E. Hanson, J.W. Watson, J.S. Lee, Drug Metab. Dispos. 31 (2003) 312e
318.
disappearance of study substrate was determined by comparing the
LC-MSpeakarea in theappropriate 0 minsample (withoutNADPH) to
peak area of the corresponding metabolized sample, by using ion
chromatograms createdwith13C-isotopepeak. The metabolicprofiles
were determined by ESI-MS peak areas of molecular ion of a partic-
ular metabolite, assuming their responses to be directly comparable.
[12] B. Brudeli, L.R. Moltzau, K.W. Andressen, K.A. Krobert, J. Klaveness, F.O. Levy,
Bioorg. Med. Chem. 18 (2010) 8600e8613.
[13] M.A. Alisi, N. Cazzolla, R. Costi, R. Di Santo, G. Furlotti, A. Guglielmotti,
L. Polenzani, PCT Int. Appl. (July 29, 2008)WO2010/012611A1.
[14] M. Fedouloff, F. Hossner, M. Voyle, J. Ranson, J. Powles, G. Riley, G. Sanger,
Bioorg. Med. Chem. 9 (2001) 2119e2128.
[15] L.M. Gaster, F.D. King, Med. Res. Rev. 17 (1997) 163e214.
[16] M.L. López-Rodríguez, B. Benhamú, A. Viso, M.J. Morcillo, M. Murcia,
L. Orensanz, M.J. Alfaro, M.I. Martın, Bioorg. Med. Chem. 7 (1999) 2271e
2281.
Conflict of interest
The compounds described in the present paper are described in
WO2010112865 (Klaveness, J.; Brudeli, B.; Levy, F. O.; Moltzau, L. R.;
Gulbrandsen, T.). This patent family is owned by Serodus AS where
Andressen, Klaveness, Levy and Moltzau are shareholders.
[17] H.J. Witchel, Cardiovasc. Ther. 29 (2011) 251e259.
[18] L. Kiss, P.B. Bennett, V.N. Uebele, K.S. Koblan, S.A. Kane, B. Neagle,
K. Schroeder, Assay Drug Dev. Technol. 1 (2003) 127e135.
[19] R. Waterhouse, Mol. Imaging Biol. 5 (2003) 376e389.
[20] S.D. Krämer, N.J. Abbott, D.J.A. Begley, Biological models to study blood-brain
barrier permeation, in: B. Testa, H. Waterbeemd, G. Folkers, R.H. Guy (Eds.),
Pharmacokinetic Optimization in Drug Research: Biological, Physiochemical
and Computational Strategies, Wiley-VHCA, Zürich, 2001, pp. 127e153.
[21] X. Liu, B. Testa, A. Fahr, Pharm. Res. 28 (2011) 962e977.
[22] V.A. Levin, J. Med. Chem. 23 (1980) 682e684.
Acknowledgments
This work was supported by The Research Council of Norway
and Serodus AS. The authors wish to thank Trygve Gulbrandsen,
Ph.D., COO of Serodus AS, for support, advice and discussions.
References
[23] W.M. Pardridge, D. Triguero, J. Yang, P.A. Cancilla, J. Pharmacol. Exp. Ther. 253
(1990) 884e891.
[24] H. Ansanay, A. Dumuis, M. Sebben, J. Bockaert, L. Fagni, Proc. Natl. Acad. Sci. U.
S. A. 92 (1995) 6635e6639.
[25] G.E. Torres, Y. Chaput, R. Andrade, Mol. Pharmacol. 47 (1995) 191e197.
[26] H. Mei, W. Korfmacher, R. Morrison, AAPS J. 8 (2006) E493eE450.
[27] B. Davies, T. Morris, Pharm. Res. 10 (1993) 1093e1095.
[28] M. Langlois, R. Fischmeister, J. Med. Chem. 46 (2003) 319e344.
[29] P. Ettmayer, G.L. Amidon, B. Clement, B. Testa, J. Med. Chem. 47 (2004) 2393e
2404.
[30] L.J. Scott, P.L. McCormack, Drugs 68 (2008) 1239e1272.
[31] D.E. Mire, T.N. Silfani, M.K. Pugsley, J. Cardiovasc. Pharmacol. 46 (2005) 583e
593.
[1] J. Hannon, D. Hoyer, Behav. Brain Res. 195 (2008) 198e213.
[2] D.K. Arulmozhi, A. Veeranjaneyulu, S.L. Bodhankar, Vasc. Pharmacol. 43 (2005)
176e187.
[3] A.B. Kaumann, F.O. Levy, Pharmacol. Ther. 111 (2006) 674e706.
[4] E. Qvigstad, T. Brattelid, I. Sjaastad, K.W. Andressen, K.A. Krobert,
J.A. Birkeland, O.M. Sejersted, A.J. Kaumann, T. Skomedal, J.-B. Osnes, F.O. Levy,
Cardiovasc. Res. 65 (2005) 869e878.
[5] J.A.K. Birkeland, I. Sjaastad, T. Brattelid, E. Qvigstad, E.R. Moberg, K.A. Krobert,
R. Bjørnerheim, T. Skomedal, O.M. Sejersted, J.-B. Osnes, F.O. Levy, Br. J.
Pharmacol. 150 (2007) 143e152.
[6] J.K. Kjekshus, C. Torp-Pedersen, L. Gullestad, L. Køber, T. Edvardsen, I.C. Olsen,
I. Sjaastad, E. Qvigstad, T. Skomedal, J.-B. Osnes, F.O. Levy, Eur. J. Heart Failure
11 (2009) 771e778.
[32] P. Laeis, K. Puchler, W. Kirch, J. Hypertens. (Suppl. 19) (2001) S21eS32.
[7] A. Sikander, S.V. Rana, K.K. Prasad, Clin. Chim. Acta 403 (2009) 47e55.
́