Bis-tetrahydroisoquinoline derivatives: AG525E1, a new step in the search for non-quaternary non-peptidic small conductance Ca2+-activated K+ channel blockers

Article abstract of DOI:10.1016/j.bmcl.2008.03.069

So far, small conductance Ca²⁺-activated K⁺ channel (SK) blockers mostly consist of quaternary ammonium derivatives or peptides. Due to their physicochemical properties, these blockers are not suitable to study the physiological role

Full text of DOI:10.1016/j.bmcl.2008.03.069

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3440–3445
Bis-tetrahydroisoquinoline derivatives: AG525E1, a new step
in the search for non-quaternary non-peptidic small conductance
Ca2+-activated K+ channel blockers
a
b
b
c
c
´
´
Amaury Graulich, Cedric Lamy, Livia Alleva, Sebastien Dilly, Philippe Chavatte,
d
b
a,
*
´
Johan Wouters, Vincent Seutin and Jean-Franc¸ois Liegeois
a
` ˆ
Drug Research Center, Laboratory of Medicinal Chemistry, University of Liege, avenue de l’Hopital,
`
1 (B36), B-4000 Liege 1, Belgium
bResearch Center for Cellular and Molecular Neurobiology, Laboratory of Pharmacology,
`
ˆ
`
University of Liege, avenue de l’Hopital, 1 (B36), B-4000 Liege 1, Belgium
cFaculte´ des Sciences Pharmaceutiques et Biologiques, Laboratoire de Chimie The´rapeutique,
EA 1043, Universite´ de Lille 2, rue du Professeur Laguesse, 3, BP83, F-59006 Lille Cedex, France
dDepartment of Chemistry, University of Namur, rue de Bruxelles, 61, B-5000 Namur, Belgium
Received 18 February 2008; revised 21 March 2008; accepted 25 March 2008
Available online 29 March 2008
Abstract—So far, small conductance Ca2+-activated K+ channel (SK) blockers mostly consist of quaternary ammonium derivatives
or peptides. Due to their physicochemical properties, these blockers are not suitable to study the physiological roles of SK channels
in the central nervous system in vivo. Herein, we report the discovery of a chiral bis-tertiary amine with SK blocking properties from
chemical modulation of laudanosine. AG525E1 has an affinity for SK channels (Ki = 293 nM) approximately 100-fold higher than
the tertiary compound laudanosine (Ki ꢀ 30 lM) and similar to the charged compound dequalinium (Ki = 221 nM). AG525E1 equi-
potently blocks SK1, SK2 and SK3 currents in transfected cell lines. Because of its basic and lipophilic properties, it can reach cen-
tral SK targets.
ꢀ 2008 Elsevier Ltd. All rights reserved.
Small conductance Ca2+-activated K+ (SK) channels are
found in several types of neurons as well as in other cell
types. The SK channels are selective for K+ ions and
their opening depends upon an increase of the intracel-
lular Ca2+ concentration. These channels underlie the
medium postspike afterhyperpolarization (mAHP),
which plays an important role in modulating the firing
rate and the firing pattern of neurons.1,2 Three SK chan-
nel subtypes have been cloned1 and exhibit a differential
location in the brain. The distribution of the SK channel
subtypes was investigated in the rat by using in situ
hybridization and immunohistochemistry and revealed
that SK1 and SK2 subtypes are mostly expressed in
the cortex and hippocampus3 whilst SK3 subtype
expression is higher in the monoaminergic cell regions.
These features attract great attention to SK channels
as putative targets for the treatment of various CNS dis-
eases such as cognitive dysfunction,4–8 neuronal hyper-
excitability,9 dopamine-related disorders,10–12 and
depression.7
MeO
Cl -
H
N
MeO
Me
S
H
Cl -
Me
N
H
S
H
Keywords: Apamin; Tetrahydroisoquinoline; Patch-clamp; Neurons;
Motor activity; Stereoisomer.
MeO
*
Corresponding author. Tel.: +32 (0) 43 66 43 77; fax: +32 (0) 43 66 43
62; e-mail: JF.Liegeois@ulg.ac.be
AG525E1
OMe
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.069

A. Graulich et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3440–3445
3441
MeO
MeO
O
I
I
N
N
O
Me
Me
Me
Me
OMe
O
NML
NMN
OMe
OMe
O
OMe
OMe
MeO
MeO
MeO
MeO
N
N
Me
Me
OMe
8-MeO-Laudanosine
Laudanosine
OMe
OMe
OMe
OMe
SK channel blockers consist of peptidic toxins such as
apamin,13 leiurotoxin I,14 PO5,15 tityus j,16 tamapin,17
BmSKTx1,18 and non-peptidic compounds such as
tubocurarine,19 atracurium,20 dequalinium21 and UCL
derivatives.22,23 We reported that N-methyl-laudanosine
(NML)24–26 and N-methyl-noscapine (NMN)26 also
block SK channels. Unlike apamin, these two molecules
are medium affinity blockers with a high reversibility as
shown in patch-clamp experiments.26 Besides their low
affinities, the quaternary ammonium function is proba-
bly the major drawback of these pharmacological tools
for investigating their CNS effects after systemic injec-
tion. Indeed, this permanent positive charge imposes
to inject the drugs by an invasive route, such as the
intracerebroventricular one.27 Some tertiary derivatives
have also been reported to block SK channels28–30 and
in the first part of our program 8-MeO-laudanosine
was found to be almost equipotent to NML at interact-
ing with apamin-sensitive sites with a Ki of 5.8 lM ver-
sus 1.6 lM for the latter.31 Previous studies had shown
that the presence of two positive charges or two basic
centres improves the affinity of the blockers.32–34 In
order to find new compounds with significant affinity
and appropriate physicochemical properties, bis-tertiary
amines based on the scaffold of laudanosine were pre-
pared and tested for their affinity on SK channels.35
We report herein the results obtained with the most
potent basic compound in this series.
The bis-tertiary amine was obtained from 6,7-dime-
thoxy-isoquinoline which was classically prepared using
a modification of the Pomeranz–Fritsch synthesis
(Scheme 1).36 The bis-alkylation was subsequently per-
formed by using the Reissert compound pathway
Me
Me
N
(CH )
NH
N
H N
2
2
2 10
Dequalinium
NH
N
NH
N
NH
NH
N
N
UCL1684
UCL1848

3442
A. Graulich et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3440–3445
MeO
MeO
MeO
MeO
a
N
N
Bz
1
H
CN
b, c
MeO
MeO
MeO
I
N
Me
N
MeO
d
(CH )
2 3
(CH )
2 3
Me
OMe
N
OMe
OMe
N
I
OMe
3
2
MeO
MeO
e
N
Me
(CH )
2 3
Me
OMe
OMe
N
4
Scheme 1. Reagents and conditions: (a) Me3SiCN, BzCl, AlCl3, CH2Cl2, rt, 70%; (b) I-(CH2)3-I, NaH, DMF, ꢁ10 ꢁC; (c) NaOH, EtOH/H2O, reflux,
31%; (d) MeI, DMF, Dt until dissolution, 97%; (e) NaBH4, MeOH, rt, 67%.
ate bis-electrophilic reagent.39 The alkylated Reissert
compound was hydrolysed to give the bis-isoquinoline
derivative (2). The bis-isoquinoline dissolved in DMF
with a mild heating was methylated by methyl iodide
to obtain the corresponding bis-isoquinolinium deriva-
tive (3).34 Finally, the mixture of bis-tetrahydroisoquin-
oline stereoisomers (4) was obtained by the reduction of
the bis-isoquinolinium analogue with an excess of so-
dium borohydride (Scheme 1).40 After biological screen-
ing at 10 lM, the three stereoisomers were separated by
semi-preparative chiral HPLC.41 AG525E1 was the first
eluted enantiomer and was subsequently isolated as a
dihydrochloride. In our binding conditions as previously
reported (see 25,34,42), NML and dequalinium (DQ+)
Figure 1. SK3 current traces recorded at a holding potential of
had an affinity (Ki) for the apamin-sensitive sites of
ꢁ50 mV during 1 s followed by a 1 s ramp from ꢁ50 mV to +50 mV.
ꢀ1600 nM and ꢀ220 nM, respectively.30 The affinities
The current was blocked by increasing concentration of AG525E1.
of the three stereoisomers were 293 22 nM,
Their effect is represented by different grey levels. From the bottom to
1422 116 nM and 1885 105 nM for the first eluted
(AG525E1), the meso form and the second eluted iso-
the top trace, the concentrations used were 0.3, 1, 3, 5, 10, and 30 lM.
NML was used at a supra-maximal concentration as a reference.
mer, respectively. The difference in affinity between the
two enantiomers (AG525E1: Ki = 293 22 nM; second
enantiomer: Ki = 1885 105 nM) indicates that the ste-
reochemistry is an important feature for interacting with
SK channels, while enantiomers of NML were previ-
ously found to be equipotent on apamin-sensitive sites.25
The absolute configuration of AG525E1 was determined
by X-ray crystallography after the crystallisation of the
dihydrochloride from MeCN.43 This experiment showed
(Scheme 1).37 The Reissert compound was obtained by
the reaction of the corresponding isoquinoline with ben-
zoyl chloride in the presence of trimethylsilyl cyanide in
a good yield.38 This derivative was deprotonated by so-
dium hydride in DMF and the resulting Reissert anion
was alkylated by using a half equivalent of the appropri-

A. Graulich et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3440–3445
3443
that it possesses a S,S configuration associated with a
dextrorotatory activity. The affinity of this enantiomer
(Ki = 293 nM) is similar to that of the reference charged
compound dequalinium (Ki = 221 nM) but clearly supe-
rior to that of laudanosine (Ki ꢀ 30 lM.24 As is the case
for quaternary compounds,34 the development of non-
quaternary bis-derivatives is highly favourable for the
interaction with SK channels. From laudanosine, an
approximately 100-fold increase in affinity is observed.
An extensive binding profile performed in another labo-
ratory on 65 receptors or channels and 7 enzymes at
10 lM confirms that the compound has a high affinity
for SK channels (61% inhibition of specific binding) ver-
sus other targets. AG525E1 has weaker affinity (% inhi-
bition of specific binding) for serotonin 5-HT1A (39%),
sigma2 (39%), histamine H3 (37%) and dopamine D2
receptors (31%) and the verapamil site of the L type cal-
cium channel (36%).
nosine. We demonstrate the feasibility of targeting cen-
tral SK channels after the systemic administration of
non-quaternized compounds. Further efforts will be di-
rected towards the discovery of molecules with a higher
selectivity for SK channels versus other targets, on the
one hand, and subtype-specific blockers, on the other
hand. Although the search for high affinity ligands is fre-
quently a challenge in medicinal chemistry, medium
affinity blockers with a high reversibility, as shown in
patch-clamp experiments for NML,26 will be valuable
tools for electrophysiological exploration of SK channel
physiology.
Acknowledgments
The technical assistance of Y. Abrassart, S. Counerotte,
and J.-C. Van Heugen is gratefully acknowledged. A.G.
was supported by a Research Fellow grant of the
Whole-cell patch-clamp experiments in transfected cell
lines were used to confirm that the compound effectively
blocks SK currents (Fig. 1).44 AG525E1 completely
blocked SK1, SK2 and SK3 currents with IC50’s of
3.8 0.8 lM (n = 13), 2.6 0.3 lM (n = 13) and
3.0 0.4 lM (n = 14), respectively.
`
«Fonds pour la formation a la Recherche Industrielle
et Agricole (F.R.I.A.)» and J.-F.L. is a Senior Research
Associate of the ‘Fonds National de la Recherche Scien-
tifique’ (F.N.R.S.) of Belgium. C.L. is the recipient of a
‘First DEI’ Ph.D. Studentship from the Walloon Region
of Belgium (Contract No. 516131). This work was finan-
cially supported in part by F.N.R.S. Grant Nos.
3.4525.98 and 9.4560.03 and by grants from Fonds
The affinity for the apamin-sensitive binding sites of SK
channels depends on the presence of positive charges in
the structure,32 but, in order to develop compounds with
potential effect on CNS, the basic character of the mole-
cule is important in terms of bioavailability. Unlike
NML, dequalinium and UCL1684, the bis-tertiary amine
AG525E1 does not bear one or two permanent positive
charge(s). Therefore, physicochemical parameters such
as pKa and logP were determined using potentiometric
titration procedures on a Siriusꢂ PCA200 apparatus.45
AG525E1 possesses two values of pKa, namely 9.3 and
8.1. Therefore, at physiological pH, this stereoisomer
has mostly (ꢀ83%) a positive charge on both nitrogens,
and a small fraction (ꢀ17%) is ionised just once whilst
the free base represents a minor proportion (<1%). The
lipophilicity of the compound was also measured by
potentiometric titration using the same equipment.45
The logP value is 2.90. More generally from a theoretical
point of view, the structure of AG525E1 is in accordance
with the classical Lipinski’s Rule of Five. Indeed, this ste-
reoisomer possesses six H bond acceptors and has a logP
which is smaller than 5. It has a molecular weight of ꢀ527
daltons and does not possess H bond donors.46 All these
parameters support the ability of the compound to reach
CNS targets after crossing the blood–brain barrier. Preli-
minary experiments in rats showed that, following ip
injection, the compound tends to increase motor activity
(data not shown) as previously reported for apamin.10
Of course, this effect is quite unspecific and the compound
will be better characterized in further pharmacological
and psychopharmacological experiments since the non-
selectivity for SK channel subtypes probably leads to a
global behaviour resulting from combined SK channel
subtype blockade.
´
Speciaux pour la Recherche 2002 and 2003 of the Uni-
versity of Liege and the Fondation Leon Fredericq of
`
the Faculty of Medicine of the University of Liege.
`
´
´ ´
References and notes
1. Kohler, M.; Hirschberg, B.; Bond, C. T.; Kinzie, J. M.;
Marrion, N. V.; Maylie, J.; Adelman, J. P. Science 1996,
273, 1709.
2. Stocker, M.; Krause, M.; Pedarzani, P. Proc. Natl. Acad.
Sci. U.S.A. 1999, 96, 4662.
3. Stocker, M.; Pedarzani, P. Mol. Cell Neurosci. 2000, 15,
476.
4. Messier, C.; Mourre, C.; Bontempi, B.; Sif, J.; Lazdunski,
M.; Destrade, C. Brain Res. 1991, 551, 322.
5. Deschaux, O.; Bizot, J. C.; Goyffon, M. Neurosci. Lett.
1997, 222, 159.
6. Ikonen, S.; Riekkinen, P., Jr. Eur. J. Pharmacol. 1999, 382,
151.
7. Van der Staay, F. J.; Fanelli, R. J.; Blokland, A.; Schmidt,
B. H. Neurosci. Biobehav. Rev. 1999, 23, 1087.
8. Fournier, C.; Kourrich, S.; Soumireu-Mourat, B.; Mour-
re, C. Behav. Brain Res. 2001, 121, 81.
9. Pedarzani, P.; Mosbacher, J.; Rivard, A.; Cingolani, L. A.;
Oliver, D.; Stocker, M.; Adelman, J. P.; Fakler, B. J. Biol.
Chem. 2001, 276, 9762.
10. Steketee, J. D.; Kalivas, P. W. J. Pharmacol. Exp. Ther.
1990, 254, 711.
11. Shepard, P. D.; Bunney, B. S. Exp. Brain Res. 1991, 86,
141.
12. Seutin, V.; Johnson, S. V.; North, R. A. Brain Res. 1993,
630, 341.
13. Habermann, E.; Fischer, K. Eur. J. Biochem. 1979, 94,
355.
14. Castle, N.; Strong, P. FEBS Lett. 1986, 209, 117.
15. Zerrouk, H.; Mansuelle, P.; Benslimane, A.; Rochat, H.;
Martin-Eauclaire, M. FEBS Lett. 1993, 320, 189.
In summary, a new scaffold of tertiary SK channel
blockers has been obtained from the structure of lauda-

3444
A. Graulich et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3440–3445
16. Legros, C.; Oughedeni, R.; Darbon, H.; Rochat, H.;
Bougis, P.; Martin-Eauclaire, M.-F. FEBS Lett. 1996, 390,
81.
17. Pedarzani, P.; D’hoedt, D.; Doorty, K.; Wadsworth, J.;
Joseph, J.; Jeyaseelan, K.; Manjunatha Kini, R.; Gadre,
S.; Sapanekar, S.; Stocker, M.; Strong, P. J. Biol. Chem.
2002, 277, 46101.
18. Xu, C.-Q.; He, L.-L.; Brøne, B.; Martin-Eauclaire, M.-F.;
Van Kerkhove, E.; Zhou, Z.; Chi, C.-W. Toxicon 2004, 43,
961.
19. Nohmi, M.; Kuba, K. Brain Res. 1984, 301, 146.
20. Cook, N.; Haylett, D. J. Physiol. 1985, 358, 373.
21. Castle, N.; Haylett, D.; Morgan, J.; Jenkinson, D. Eur. J.
Pharmacol. 1993, 236, 201.
with a mixture of Me2CO/MeOH (9:1, v/v) as mobile
phase. The purified oil was isolated as a hydrochloride salt
and further recrystallized from MeCN/Et2O; yield, 67%;
m/z 455 (100) [MH+], 146 (11) [THIQH+]; Anal.
(C27H38N2O4Æ2HClÆ3/4H2O) C, H, N.
41. The semi-preparative resolution was performed by HPLC
using a liquid chromatograph consisting of a pump Armen
Instrument AP100 with an injection valve Knauerꢂ (1–
10 mL). The stationary phase was a Chiralcelꢂ OD-H
column (20 · 250 mm) preceded by a Chiralcelꢂ OD-H
precolumn (10 · 20 mm), while the mobile phase was a
mixture of n-hexane/2-propanol (8:2, v/v) +0.05% DEA.
Samples were collected by using a Buchi Fraction-Collec-
¨
tor C-660 every 45 s. The fractions containing AG525E1
(rt: 7.3 min) were pooled and evaporated. The residue was
kept in a desiccation unit under vacuum during 48 h. The
purity (97.4%) and the diastereoisomeric excess (99.7%)
were determined by analytical chiral HPLC using a
Chiralcelꢂ OD-H column (4.6 · 250 mm) with the same
mobile phase. After this control, the solid residue was
dissolved in Et2O (5–10 mL) and the dihydrochloride was
precipitated by the addition of saturated etherous HCl
solution. The white precipitate was collected and kept
under a dry atmosphere until biological testing. A sample
was recrystallized from a mixture of MeOH and Et2O, and
22. Galanakis, D.; Ganellin, R.; Malik, S.; Dunn, P. J. Med.
Chem. 1996, 39, 3592.
23. Campos Rosa, J.; Galanakis, D.; Piergentili, A.; Bhandari,
K.; Ganellin, C. R.; Dunn, P. M.; Jenkinson, D. H. J.
Med. Chem. 2000, 43, 420.
´
´
24. Scuvee-Moreau, J.; Liegeois, J.-F.; Massotte, L.; Seutin,
V. J. Pharmacol. Exp. Ther. 2002, 302, 1176.
´
25. Graulich, A.; Mercier, F.; Scuvee-Moreau, J.; Seutin, V.;
´
Liegeois, J.-F. Bioorg. Med. Chem. 2005, 13, 1201.
´
26. Scuvee-Moreau, J.; Boland, A.; Graulich, A.; Van Over-
meire, L.; D’hoedt, D.; Graulich-Lorge, F.; Thomas, E.;
´
Abras, A.; Stocker, M.; Liegeois, J.-F.; Seutin, V. Br. J.
Pharmacol. 2004, 143, 753.
27. Waroux, O.; Masotte, L.; Alleva, L.; Graulich, A.;
1
analyzed as a dihydrochloride except for H NMR. mp:
20
257–258 ꢁC, dec.; ½aꢂD : +35ꢁ (MeOH); IR (KBr, cmꢁ1):
2936, 2542, 1614, 1522, 1230, 1114, 1000. 1H NMR
(CDCl3) d 1.50–1.55 (m, 2H), 1.65–1.70 (m, 2H), 1.72–1.78
(m, 2H), 2.43 (s, 6H), 2.59 (dt, 2H, J = 5.2 and 16.0 Hz),
2.69 (dt, 2H, J = 5.2 and 12.4 Hz), 2.74–2.80 (m, 2H), 3.12
(dq, 2H, J = 5.2 and 12.4 Hz), 3.36 (t, 2H, J = 5.6 Hz),
3.81 (s, 6H), 3.84 (s, 6H), 6.54 (s, 4H).
´
´
Thomas, E.; Liegeois, J.-F.; Scuvee-Moreau, J.; Seutin,
V. Eur. J. Neurosci. 2005, 22, 3111.
28. Zunszain, P. A.; Shah, M. M.; Miscony, Z.; Javadzadeh-
Tabatabaie, M.; Haylett, D. G.; Ganellin, C. R. Arch.
Pharm. (Weinheim) 2002, 4, 159.
´
42. Graulich, A.; Scuvee-Moreau, J.; Alleva, L.; Lamy, C.;
´
Waroux, O.; Seutin, V.; Liegeois, J.-F. J. Med. Chem.
29. Shah, M. M.; Javadzadeh-Tabatabaie, M.; Benton, D. C.
H.; Ganellin, C. R.; Haylett, D. G. Mol. Pharmacol. 2006,
70, 1494.
30. Fletcher, D. I.; Ganellin, C. R.; Piergentili, A.; Dunn, P.
M.; Jenkinson, D. H. Bioorg. Med. Chem. 2007, 15,
5457.
2006, 49, 7208.
43. Graulich, A.; Norberg, B.; Liegeois, J.-F.; Wouters J. Acta
Cryst E., in press.
´
44. Whole-cell patch-clamp experiments were performed in a
voltage clamp mode using an EPC-10 amplifier (Heka)
and Patchmaster software (Heka). Borosilicate pipettes
were pulled with a P87 puller (Sutter) and fire-polished
using a Microforge (Narishige) to achieve a tip resistance
between 2 and 3 MX. All the experiments with series
resistances above 8 MX were discarded. An eight channel
perfusion system (Ala science) was used for drug applica-
tions. The composition of the external solution was (in
mM): KCl 144, CaCl2 2, MgCl2 1, Hepes 10, glucose 10.
The pH was adjusted to 7.4 using KOH and the
osmolarity ranged from 290 to 305 mOsm. The internal
solution consisted of (in mM): KCl 130, EGTA 10, CaCl2
8.751, MgCl21.08, Hepes 10 leading to 1 mM free calcium.
The pH was adjusted to 7.2 using KOH and the
osmolarity ranged from 275 to 290 mOsm. Experiments
were achieved with stably transfected CHO FlipIn cells
(SK1, 2 or 3) which were kept under standard culture
conditions (medium composition was Ham’s F-12 with
10% calf serum supplemented with streptomycin/penicillin
and hygromycin-B). Potassium currents were measured at
ꢁ50 mV. A supra-maximal concentration of N-methyl-
laudanosine (100 lM) was used to achieve a complete
inhibition of the SK current and allowed to calculate the
percentage of inhibition observed with the increasing
concentrations of AG525E1.
´
´
31. Graulich, A.; Scuvee-Moreau, J.; Seutin, V.; Liegeois, J.-
F. J. Med. Chem. 2005, 48, 4972.
32. Galanakis, D.; Ganellin, R. Bioorg. Med. Chem. Lett.
2004, 14, 4031.
33. Galanakis, D.; Ganellin, R.; Dunn, M.; Jenkinson, D.
Arch. Pharm. (Weinheim) 1996, 326, 524.
´
34. Graulich, A.; Dilly, S.; Farce, A.; Scuvee-Moreau, J.;
Waroux, O.; Lamy, C.; Chavatte, P.; Seutin, V.; Liegeois,
´
J.-F. J. Med. Chem. 2007, 50, 5070.
´
´
35. Graulich, A.; Liegeois, J.-F.; Scuvee-Moreau, J.; Seutin,
V. PCT WO2007107442-A2, 2007.
36. Birch, A. J.; Jackson, A. H.; Shannon, V. R. J. Chem. Soc.
Perkin I 1974, 2185–2190.
37. Popp, F. D. Adv. Heterocycl. Chem. 1979, 24, 187.
38. Ruchirawat, S.; Phadungkul, N.; Chuankamnerdkarn, M.;
Thebtaranonth, C. Heterocycles 1977, 6, 43.
39. Popp, F.; Klinowski, C.; Piccirilli, R.; Purcell, D.; Watts,
R. J. Heterocycl. Chem. 1971, 8, 313.
40. Under an inert atmosphere, NaBH4 (1.76 g; 46.5 mmol)
was added to a solution of the 1,3-bis[1-(6,7-dimethoxy-2-
methylisoquinolylium)]-propane
diiodide34
(1.85 g;
3.1 mmol) in MeOH (100 mL) at room temperature. After
15 min, MeOH was removed under reduced pressure and
the crude residue was dissolved in a 1 N aqueous HCl
(100 mL). The acidic layer was washed with Et2O (3·
20 mL), and then basified with NH4OH. The suspension
was extracted with CH2Cl2 (3· 30 mL). The organic layers
were collected, dried over anhydrous MgSO4 and evapo-
rated under reduced pressure to afford a colourless oil,
which was purified by flash chromatography on Kieselgel
45. pKa and logP values were determined using the Sirius
PCA200TM analyser based on a potentiometric titration
method according to the previously described meth-
ods.47,48 Potentiometric pKa values were determined by
titrating a minimum of three aqueous solutions of the

A. Graulich et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3440–3445
3445
samples containing 32–56% (weight%) methanol. In each
experiment, 10 mL of 0.5–1 mM solution of the
from 15–0.5 to 7.5–10 for the proportions of water–
octanol, respectively. From the octanol-containing titra-
tions, the poKa and then logP values were estimated and
refined by the RefinementProTM software, where the
aqueous pKa values previously determined were used as
unrefined contributions. A minimum of three titrations at
different phase volume ratios were made, and the mean
logP was calculated. The relevant relationships between
logP, pKa, and poKa are previously described.47.
a
sample was acidified to pH 1.8–2.0 with 0.5 M aqueous
HCl, and then titrated with 0.5 M aqueous KOH to pH
12. The titrations were carried out at constant ionic
strength (I = 0.15 M KCl) and temperature (T = 25.0
0.5 ꢁC). Then, these psKa values (the apparent ionisa-
tion constants in aqueous methanolic solution) were
used to estimate the aqueous pKa values. There after,
potentiometric logP determinations were carried out.
Typically, 7.5–17.5 mL of 0.5–1 mM solutions of sam-
ples was titrated under the same conditions as above but
in the presence of various proportions of the partition-
ing solvent namely water-saturated octanol. Depending
on the expected logP, the phase volume ratio varied
46. Lipinski, C.; Lombardo, F.; Dominy, B.; Feeney, P. Adv.
Drug Deliv. Rev. 1997, 23, 3.
47. Avdeef, A. J. Pharm. Sci. 1993, 82, 183.
48. Avdeef, A.; Box, K. J.; Comer, J. E. A.; Gilges, M.;
Hadley, M.; Hibbert, C.; Patterson, W.; Tam, K. Y. J.
Pharm. Biomed. Anal. 1999, 20, 631.