J. Wilkinson et al. / European Journal of Pharmacology 561 (2007) 160–163
161
Fig. 1. Synthesis of the 2′-hydroxy analogues of fendiline. Conditions: a. i. 2.2 equivalents sec-BuLi, Et2O, 0 °C, 5 h. ii. allyl bromide, rt, 18 h, 85%. b. Ozone,
CH2Cl2, −78 °C, 15 min, 19%. c. S-Phenethylamine, NaCNBH3, MeOH, HCl, rt, 30 h. d. R-Phenethylamine, NaCNBH3, MeOH, HCl, rt, 30 h, 33%.
2004). Initially a dilithium species was formed from 4 and
quenching with allyl bromide gave the racemic allyl compound
5 (Wilkinson et al., 2004a). This was subjected to ozonolytic
cleavage to give, 4-phenyl-chroman-2-ol, 6, as a diastereomeric
mixture of hemiacetals. Reaction with S-phenethylamine under
reductive amination conditions gave a 1:1 mixture of the pair of
diastereomers (R, S)-2-[1-phenyl-3-(1-phenyl-ethylamino)-pro-
pyl]-phenol, 2, and (S, S)-2-[1-phenyl-3-(1-phenyl-ethyla-
mino)-propyl]-phenol, 3 which were cleanly isolated by
column chromatography. Tentative assignment of stereochem-
istry at the newly created stereocentre was achieved by
comparison of spectral data with that from an enantioenriched
product. Alkene 5 was synthesised with the chiral ligand (−)-
sparteine present in 46% ee with an S-configuration for the
major product (assigned by conversion to a known compound,
see Wilkinson et al., 2006). Subsequently 6 was prepared with a
predominant S-configuration at the newly created stereocentre
(46% de) and reacted with 2S-phenylethylamine. The major
“run up” procedure consisting of 2×2 min exposures to 60 mM
KCl (isosmotically substituted for NaCl in PSS). Tissues were
once again contracted up with high K+ solution (60 mM) and,
once a stable level of contraction had been achieved, cumulative
concentration-response curves were constructed to the two
isomers of fendiline (R-1 and S-1) and the substituted derivates
(compounds 2 and 3). All responses were recorded on a chart
recorder for later analysis. The presence of a functional
endothelium was tested for by observing responses to carbachol
(50 μM).
Responses were expressed as changes in active wall tension
(mN/mm) from resting levels and normalized, as appropriate, to
responses to 60 mM KCl. For comparison of sensitivity all
curves were normalized to the maximal response obtained at a
concentration of 100 μM and EC50 values were calculated
using Prism. Statistical significance in the two groups of arteries
was examined by performing one-way ANOVA followed by
Newman Keuls Multiple Comparison Test on logEC50 values.
Responses of mesenteric and coronary arteries to similar
compounds were compared by unpaired Student's t test.
All protocols were approved the local institutional ethics
committee.
1
diastereomer formed had identical H and 13C NMR spectra to
compound 3 which was therefore assigned as the S, S
diastereomer.
2.2. Tissue preparation and measurements of contractility
3. Results
Male Wistar rats (200–250 g) were killed by stunning and
cervical dislocation. From each rat a mesenteric (third-to fourth-
order branches) and coronary (first branch of septal artery)
(200–350 μm diameter) artery were dissected out and mounted
in a double chamber Mulvany–Halpern myograph for mea-
surement of isometric tension. Vessels were bathed in
physiological salt solution (PSS) of composition (mM): NaCl
119, KCl 4.7, MgSO4 ·7H2O 1.2, NaHCO3 25, KH2PO4 1.17,
EDTA 0.03, glucose 5.5, calcium 2.5, pH 7.4 at 37 °C, and were
gassed with 95% air/5% CO2. After mounting and equilibration,
the resting tension to internal circumference ratio of arteries was
determined and the vessels set to a normalized internal
circumference of lo where lo is 0.9 l100 and l100 is the internal
circumference of the vessel under an effective transmural
pressure of 100 mm Hg. Under these conditions, maximum
active tension is developed (Mulvany and Halpern, 1977).
Following normalization, vessels were subjected to a routine
Mesenteric and coronary arteries used in this study had mean
diameters of 255 10 μm (n=24) and 239 8 μm (n=23)
respectively. Addition of 60 mM KCl resulted in an initial rapid
contraction which relaxed slightly to a stable level. The mean
magnitude of this maintained contraction was 2.83 0.18 mN/
mm for mesenteric arteries and 1.73 0.14 mN/mm for coronary
arteries. This contraction was maintained (in control experi-
ments) for the duration of the experimental protocol (approx.
100 min). Mean maximal relaxation to the endothelium-
dependent dilator carbachol was 66 24% KCl contraction.
All compounds relaxed both mesenteric and coronary pre-
constricted arteries (to baseline levels) in a concentration-
dependent manner (Fig. 2), however, there were significant
differences in the sensitivity of arteries to the different
compounds. Both mesenteric and coronary arteries were
significantly more sensitive (as determined by comparison of