Sila-trifluperidol, a silicon analogue of the dopamine (D2) receptor antagonist trifluperidol: Synthesis and pharmacological characterization

Article abstract of DOI:10.1021/om901011t

Trifluperidol (2a) is a dopamine (D₂) receptor antagonist of the butyrophenone type. Carbon/silicon exchange (sila-substitution) in the 4-position of the piperidine ring of 2a (R₃COH → R ₃SiOH) results in sila-trifluperidol (2b). The silanol 2b was synthesized in a multistep synthesis, starting from triethoxy(vinyl)silane, and was isolated as the hydrochloride 2b-HCl. Compound 2b-HCl was structurally characterized by single-crystal X-ray diffraction and solution NMR spectroscopy, and the stability of the silanol 2b in aqueous solutions at different pH values was studied by ESI-MS experiments. In addition, the pharmacological profiles of the C/Si analogues trifluperidol (2a) and sila-trifluperidol (2b) and of the related compounds haloperidol (1a) and sila-haloperidol (1b) were studied in functional receptor assays at human dopamine D₁ and D₂ receptors. Sila-substitution of 1a (→ 1b) and 2a (→ 2b) affects the pharmacological properties significantly.

Full text of DOI:10.1021/om901011t

1652 Organometallics 2010, 29, 1652–1660
DOI: 10.1021/om901011t
Sila-Trifluperidol, a Silicon Analogue of the Dopamine (D2) Receptor
Antagonist Trifluperidol: Synthesis and Pharmacological
Characterization
Reinhold Tacke,*,† Binh Nguyen,† Christian Burschka,† W. Peter Lippert,†
Alexandra Hamacher,‡ Christian Urban,‡ and Matthias U. Kassack‡
†
€
€
€
Institut fu€r Anorganische Chemie, Universitat Wurzburg, Am Hubland, D-97074 Wurzburg, Germany, and
‡
€
€
€
Institut fu€r Pharmazeutische und Medizinische Chemie, Universitat Dusseldorf, Universitatsstrasse 1,
D-40225 Du€sseldorf, Germany
Received November 23, 2009
Trifluperidol (2a) is a dopamine (D2) receptor antagonist of the butyrophenone type. Carbon/
silicon exchange (sila-substitution) in the 4-position of the piperidine ring of 2a (R3COH f R3SiOH)
results in sila-trifluperidol (2b). The silanol 2b was synthesized in a multistep synthesis, starting from
triethoxy(vinyl)silane, and was isolated as the hydrochloride 2b HCl. Compound 2b HCl was
3
3
structurally characterized by single-crystal X-ray diffraction and solution NMR spectroscopy, and
the stability of the silanol 2b in aqueous solutions at different pH values was studied by ESI-MS
experiments. In addition, the pharmacological profiles of the C/Si analogues trifluperidol (2a) and
sila-trifluperidol (2b) and of the related compounds haloperidol (1a) and sila-haloperidol (1b) were
studied in functional receptor assays at human dopamine D1 and D2 receptors. Sila-substitution of 1a
(f 1b) and 2a (f 2b) affects the pharmacological properties significantly.
Introduction
These side effects (which have been correlated with neuro-
toxic effects of a pyridinium-type metabolite of 1a; in this
context, see refs 14 and 15 and references therein) present a
disadvantage in therapy and give motivation for the search
of analogues that cause less EPSs. In this context, we have
recently synthesized a silicon analogue of haloperidol (1a),
sila-haloperidol (1b),14,16,17 and have evaluated the pharma-
cological properties of the C/Si analogues 1a and 1b.14,17
Whereas in competitive binding assays at human dopamine
receptors (hD1-hD5) sila-haloperidol (1b) showed a signifi-
cantly (5-fold) higher affinity for hD2 receptors than halo-
peridol (1a), the silicon compound 1b was approximately
equipotent to its carbon analogue 1a at all other human
dopamine receptors. Furthermore, the subtype selectivity of
1b for hD2 over the other human dopamine receptors was
somewhat higher than that of 1a. As already claimed on the
basis of theoretical considerations some years ago,17 we have
recently also demonstrated that the in vitro metabolic path-
way of the C/Si analogues 1a and 1b (phase I and phase II
metabolism) is totally different.14,15 A major phase I meta-
bolite of haloperidol (1a), the pyridinium-type metabolite
that is claimed to be responsible for the neurotoxic side
effects of 1a, is not formed in microsomal incubations (rat
In the late 1950s, butyrophenones (1-phenyl-1-butanones)
have been recognized as potent neuroleptic agents and since
then have been studied extensively.1-11 The dopamine (D2)
receptor antagonists haloperidol (1a) and trifluperidol (2a)
are prominent representatives of this class of drugs. Halo-
peridol is still used clinically in the treatment of schizophre-
nia, although it may cause severe extrapyramidal side effects
(EPSs) including parkinsonism and tardive dyskinesia.12,13
*To whom correspondence should be addressed. Phone: þ49-931-31-
85250. Fax: þ49-931-888-4609. E-mail: r.tacke@uni-wuerzburg.de.
(1) Janssen, P. A. J.; van de Westeringh, C.; Jageneau, A. H. M.;
Demoen, P. J. A.; Hermans, B. K. F.; van Daele, G. H. P.; Schellekens,
K. H. L.; van der Eycken, C. A. M.; Niemegeers, C. J. E. J. Med. Pharm.
1959, 1, 281–297.
(2) Janssen, P. A. J. Arzneim.-Forsch. 1961, 11, 819–824.
(3) Gallant, D. M.; Bishop, M. P.; Timmons, E.; Steele, C. A. Curr.
Ther. Res. 1963, 5, 463–471.
(4) Pratt, J. P.; Bishop, M. P.; Gallant, D. M. Curr. Ther. Res. 1964, 6,
562–571.
(5) Janssen, P. A. J. Int. Rev. Neurobiol. 1965, 8, 221–263.
(6) Niemegeers, C. J. E.; Janssen, P. A. J. Psychopharmacologia
(Berlin) 1965, 8, 263–270.
(7) Soudijn, W.; Van Wijngaarden, I.; Allewijn, F. Eur. J. Pharmacol.
1967, 1, 47–57.
(8) Janssen, P. A. J.; Allewijn, T. N. Arzneim.-Forsch. 1969, 19, 199–
208.
(9) Lewi, P. J.; Heykants, J. J. P.; Janssen, P. A. J. Arzneim.-Forsch.
€
(14) Tacke, R.; Popp, F.; Muller, B.; Theis, B.; Burschka, C.;
Hamacher, A.; Kassack, M. U.; Schepmann, D.; Wunsch, B.; Jurva,
€
(Drug Res.) 1970, 20, 1701–1705.
U.; Wellner, E. ChemMedChem 2008, 3, 152–164.
(15) Johansson, T.; Weidolf, L.; Popp, F.; Tacke, R.; Jurva, U. Drug
Metab. Dispos. 2010, 38, 73–83.
(16) Tacke, R.; Heinrich, T. (Amedis Pharmaceuticals Ltd., U.K.)
U.K. Patent Appl. GB 2382575A (June 4, 2003).
(17) Tacke, R.; Heinrich, T.; Bertermann, R.; Burschka, C.;
(10) Givant, Y.; Shani, J.; Goldhaber, G.; Serebrenik, R.; Sulman,
F. G. Arch. Int. Pharmacodyn. 1973, 205, 317–327.
(11) Honma, T.; Sasajima, K.; Ono, K.; Kitagawa, S.; Inaba, S.;
Yamamoto, H. Arzneim.-Forsch. (Drug Res.) 1974, 24, 1248–1256.
(12) Levinson, D. F. Clin. Ther. 1991, 13, 326–352.
(13) Casey, D. E. Int. Clin. Psychopharmacol. 1995, 10, 105–114.
Hamacher, A.; Kassack, M. U. Organometallics 2004, 23, 4468–4477.
r
pubs.acs.org/Organometallics
Published on Web 03/05/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 7, 2010 1653
and human liver microsomes) with sila-haloperidol (1b).15
For sila-haloperidol, three metabolites originating from
opening of the piperidine ring were observed, a metabolism
that has not been observed for the carbon analogue halo-
peridol. In addition, one significant phase II metabolite of
haloperidol is the glucuronide of the hydroxyl group bound
to the piperidine ring, whereas the analogous metabolite was
not observed for sila-haloperidol in rat, dog, and human
hepatocytes.15
characterization of the C/Si pairs 1a/1b and 2a/2b. These
studies were performed as part of our systematic investiga-
tions on sila-substituted drugs19 (for recent reviews on
silicon-based drugs, see ref 20).
Results and Discussion
Syntheses. Sila-trifluperidol (2b) was synthesized in a
multistep synthesis, starting from triethoxy(vinyl)silane (3),
and was isolated as the hydrochloride 2b HCl (Scheme 1). In
3
this synthesis, the use of the 2,4,6-trimethoxyphenyl moiety
as a protecting group for silicon played a key role.21 Thus,
treatment of 3 with [3-(trifluoromethyl)phenyl]magnesium
bromide gave diethoxy[3-(trifluoromethyl)phenyl]vinylsil-
ane (4) (84% yield), which upon reaction with (2,4,6-tri-
methoxyphenyl)lithium afforded the corresponding (2,4,6-
trimethoxyphenyl)silane 5 (70% yield). Treatment of 5 with
vinylmagnesium chloride gave the divinylsilane 6 (92%
yield), which upon reaction with 9-borabicyclo[3.3.1]nonane
(9-BBN), followed by sequential treatment with aqueous
solutions of sodium hydroxide and hydrogen peroxide, gave
the corresponding bis(2-hydroxyethyl)silane 7 (67% yield).
Reaction of 7 with methanesulfonyl chloride, in the pre-
sence of triethylamine, afforded the corresponding bis-
[2-(methanesulfonyloxy)ethyl]silane (8), which was pure
enough to be used in the following step without further puri-
fication. Thus, treatment of 8 with 3-[2-(4-fluorophenyl)-1,3-
dioxolan-2-yl]propylamine (9) gave the cyclization product,
the 4-silapiperidine 10 (37% yield). Deprotection of 10 by
treatment with hydrochloric acid finally afforded the title
compound 2b, which was isolated as the hydrochloride
In continuation of these studies, we have now succeeded in
synthesizing sila-trifluperidol (2b), a silicon analogue of
trifluperidol18 (2a) (which also shows the above-mentioned
unfavorable EPSs), and we have studied the inhibitory
potencies of the C/Si analogues 2a and 2b at human dopa-
mine D1 and D2 receptors using a functional assay. For
reasons of comparison, compounds 1a and 1b were included
in these studies. We report here on the synthesis of 2b
2b HCl (74% yield).
3
The identities of 2b HCl, 4-7, and 10 were established by
3
elemental analyses (C, H, N) and NMR studies (1H, 13C, 19F,
29Si). Compound 2b HCl was additionally characterized by
single-crystal X-ray diffraction.
(isolated as 2b HCl), the crystal structure analyses of
3
3
2a HCl and 2b HCl, solution NMR studies of 2a HCl
3
3
3
and 2b HCl (solvent, [D6]DMSO), and the pharmacological
3
Crystal Structure Analyses. The C/Si analogues trifluperidol
hydrochloride (2a HCl) and sila-trifluperidol hydrochloride
(2b HCl) were structurally characterized by single-crystal
3
(18) (a) Wragg, W. R.; Ash, A. S. F.; Creighton, A. M. (May & Baker
Ltd., U.K.) U.K. Patent Appl. GB 948,071 (January 29, 1964).
(b) Nakatsuka, I.; Kawahara, K.; Kamada, T.; Yoshitake, A. J. Labelled
Compd. Radiopharm. 1978, 14, 133–140.
3
X-ray diffraction. The crystal data and the experimental
parameters used for these studies are summarized in Table 1;
selected bond lengths and angles are given in Table 2. The
cations of 2a HCl and 2b HCl are depicted in Figures 1 and 2,
(19) Recent original publications dealing with sila-substituted drugs:
(a) Daiss, J. O.; Burschka, C.; Mills, J. S.; Montana, J. G.; Showell,
G. A.; Fleming, I.; Gaudon, C.; Ivanova, D.; Gronemeyer, H.; Tacke, R.
Organometallics 2005, 24, 3192–3199. (b) Daiss, J. O.; Burschka, C.; Mills,
J. S.; Montana, J. G.; Showell, G. A.; Warneck, J. B. H.; Tacke, R.
Organometallics 2006, 25, 1188–1198. (c) Showell, G. A.; Barnes, M. J.;
Daiss, J. O.; Mills, J. S.; Montana, J. G.; Tacke, R.; Warneck, J. B. H. Bioorg.
Med. Chem. Lett. 2006, 16, 2555–2558. (d) Ilg, R.; Burschka, C.;
Schepmann, D.; W€unsch, B.; Tacke, R. Organometallics 2006, 25, 5396–
5408. (e) B€uttner, M. W.; Burschka, C.; Daiss, J. O.; Ivanova, D.; Rochel, N.;
Kammerer, S.; Peluso-Iltis, C.; Bindler, A.; Gaudon, C.; Germain, P.; Moras,
D.; Gronemeyer, H.; Tacke, R. ChemBioChem 2007, 8, 1688–1699.
(f) Warneck, J. B.; Cheng, F. H. M.; Barnes, M. J.; Mills, J. S.; Montana,
J. G.; Naylor, R. J.; Ngan, M.-P.; Wai, M.-K.; Daiss, J. O.; Tacke, R.; Rudd, J.
A. Toxicol. Appl. Pharmacol. 2008, 232, 369–375. (g) Lippert, W. P.;
3
3
respectively.
As shown in Figures 1 and 2, the piperidinium ring of
2a HCl and the 4-silapiperidinium ring of 2b HCl adopt a
3
3
chair conformation, with the exocyclic N-organyl group and
the 3-(trifluoromethyl)phenyl group in the equatorial posi-
tions. The bond lengths and angles of both compounds are in
the expected ranges and therefore do not require any further
discussion.
Superposition of the structures of the cations of the C/Si
analogues 2a HCl and 2b HCl (Figure 3) reveals significant
3
3
€
Burschka, C.; Gotz, K.; Kaupp, M.; Ivanova, D.; Gaudon, C.; Sato, Y.;
differences in the molecular shape. Due to the longer cova-
lent radius of the silicon atom, the 4-silapiperidinium skele-
ton of 2b HCl is more “flattened” than the piperidinium
Antony, P.; Rochel, N.; Moras, D.; Gronemeyer, H.; Tacke, R. ChemMed
Chem 2009, 4, 1143–1152. (h) Tacke, R.; M€uller, V.; B€uttner, M. W.; Lippert,
W. P.; Bertermann, R.; Daiss, J. O.; Khanwalkar, H.; Furst, A.; Gaudon, C.;
Gronemeyer, H. ChemMedChem 2009, 4, 1797–1802.
(20) Reviews: (a) Tacke, R.; Linoh, H. In The Chemistry of Organic
Silicon Compounds, Part 2; Patai, S., Rappoport, Z., Eds.; Wiley & Sons:
Chichester, 1989; pp 1143-1206. (b) Bains, W.; Tacke, R. Curr. Opin. Drug
Discovery Dev. 2003, 6, 526–543. (c) Showell, G. A.; Mills, J. S. Drug
Discovery Today 2003, 8, 551–556. (d) Mills, J. S.; Showell, G. A. Expert
Opin. Invest. Drugs 2004, 13, 1149–1157. (e) Pooni, P. K.; Showell, G. A.
Mini-Rev. Med. Chem. 2006, 6, 1169–1177. (f) Sieburth, S. McN.; Chen,
C.-A. Eur. J. Org. Chem. 2006, 311–322. (g) Gately, S.; West, R. Drug Dev.
Res. 2007, 68, 156–163. (h) Franz, A. K. Curr. Opin. Drug Discovery Dev.
2007, 10, 654–671.
3
skeleton of 2a HCl. As a consequence, the C/Si analogues
3
differ in their relative orientation of the N-organyl group
toward the hydroxy and 3-(trifluoromethyl)phenyl groups.
Analogous stereochemical features have been observed for
the related C/Si analogues 1a HCl and 1b HCl.17
3
3
€
(21) Popp, F.; Natscher, J. B.; Daiss, J. O.; Burschka, C.; Tacke, R.
Organometallics 2007, 26, 6014–6028.

1654 Organometallics, Vol. 29, No. 7, 2010
Tacke et al.
Scheme 1
˚
Table 2. Selected Bond Lengths (A) and Angles (deg) for the
Table 1. Crystal Data and Experimental Parameters for the
Crystal Structure Analyses of 2a HCl and 2b HCl
Piperidinium Skeleton of 2a HCl and the 4-Silapiperidinium
3
3
3
2b HCl
Skeleton of 2b HCl
3
2a HCl
3
3
2a HCl
3
2b HCl
3
empirical formula
formula mass, g mol-1
collection T, K
λ(Mo KR), A
cryst syst
space group (No.)
a, A
b, A
C22H24ClF4NO2
445.87
100(2)
C21H24ClF4NO2Si
461.95
193(2)
0.71073
monoclinic
P21 (4)
18.074(3)
7.0906(7)
18.197(3)
110.189(17)
2188.8(5)
4
1.402
0.281
960
0.5 ꢀ 0.3 ꢀ 0.2
4.80-58.40
-24 e h e 24,
-9 e k e 9,
-24 e l e 24
23 053
11 269
0.0383
11 269
326
C-O1
1.4292(11)
1.5276(13)
1.5327(14)
1.5370(14)
1.4998(12)
1.5014(13)
1.4988(12)
1.5233(13)
1.5223(13)
111.20(7)
105.29(8)
109.96(8)
111.03(8)
110.20(8)
109.03(7)
109.97(7)
111.08(7)
112.15(7)
111.94(8)
111.89(8)
110.39(8)
110.13(8)
Si-O1
1.639(2)
1.868(3)
1.866(3)
1.877(3)
1.509(3)
1.505(3)
1.506(3)
1.516(4)
1.522(4)
105.89(12)
113.65(12)
112.30(13)
109.73(12)
112.94(13)
102.50(12)
113.6(2)
112.3(2)
111.8(2)
114.31(17)
110.88(19)
112.3(2)
C-C1
Si-C1
0.71073
monoclinic
C2/c (15)
37.5474(19)
7.0100(4)
15.8691(9)
91.993(3)
4174.3(4)
8
1.419
0.237
1856
0.27 ꢀ 0.15 ꢀ 0.10
4.34-66.92
-57 e h e 58,
-7 e k e 10,
-24 e l e 22
76 705
C-C8
C-C11
N-C9
N-C10
N-C12
C8-C9
C10-C11
O1-C-C1
O1-C-C8
Si-C8
Si-C11
N-C9
N-C10
N-C12
C8-C9
C10-C11
O1-Si-C1
O1-Si-C8
c, A
β, deg
V, A3
Z
D(calcd), g cm-3
O1-C-C11
C1-C-C8
O1-Si-C11
C1-Si-C8
μ, mm-1
F(000)
C1-C-C11
C8-C-C11
C9-N-C10
C9-N-C12
C10-N-C12
C-C8-C9
C-C11-C10
N-C9-C8
N-C10-C11
C1-Si-C11
C8-Si-C11
C9-N-C10
C9-N-C12
C10-N-C12
Si-C8-C9
Si-C11-C10
N-C9-C8
N-C10-C11
cryst dimens, mm
2θ range, deg
index ranges
no. of collected reflns
no. of indep reflns
Rint
no. of reflns used
no. of restraints
no. of params
Sa
8096
0.0439
8096
0
277
112.0(2)
existence of two isomers of the respective cations in solution
(solvent, [D6]DMSO), with molar ratios of 12:1 (2a HCl)
and 2:1 (2b HCl), respectively (Figure 4). These different
766
0.991
0.0697/0.0000
0.0444
0.1194
0.07(6)
þ0.428/-0.278
1.098
3
weight params a/bb
R1c [I > 2σ(I)]
wR2d (all data)
absolute struct param
max./min. residual
electron density, e A-3
P
0.0600/3.3470
0.0394
0.1191
3
molar ratios indicate considerable differences in the energies
of the respective two isomers of the piperidinium and 4-
silapiperidinium skeleton. These results are in agreement
with those obtained in solution NMR studies of the related
C/Si analogues 1a HCl and 1b HCl.17 The existence of the
þ0.638/-0.442
aS = { [w(Fo2 - Fc2)2]/(n - p)}0.5; n = no. of reflections; p = no. of
parameters. bw-1=σ2(Fo2) þ (aP)2 þ bP, with P=[max(Fo2,0)þ 2Fc2]/3.
3
3
3
two isomers of 2b HCl was confirmed by 2D 1H,1H EXSY
P
P
P
P
c
R1= ||Fo|-|Fc||/ |Fo|. d wR2={ [w(Fo2-Fc2)2]/ [w(Fo2)2]}0.5
.
NMR experiments (for details, see Experimental Section).
Attempts to determine the structures of the two iso-
mers (2ar/2aβ and 2br/2bβ; Figure 4) by additional NMR
experiments (1H,1H NOESY NMR) failed. However, it is
NMR Studies. The H, 13C, 19F, and 29Si NMR spectra
of the C/Si analogues 2a HCl and 2b HCl revealed the
1
3
3

Article
Organometallics, Vol. 29, No. 7, 2010 1655
Figure 1. Molecular structure of the cation in the crystal of 2a HCl.
3
Figure 2. Molecular structure of the cation in the crystal of 2b HCl.
3
Figure 3. Superposition of the piperidinium skeleton of 2a HCl
3
(black bonds) and the 4-silapiperidinium skeleton of 2b HCl
(gray bonds). The hydrogen atoms are omitted for clarity.
3
Figure 5. ESI-MS spectra of 10 μM buffered aqueous solutions
(pH 7.4) of 2b HCl, showing the signal of the ammonium cation
3
(m/z 426). The spectra were measured 15 min (left) and 24 h
(right) after sample preparation at 20 °C (for details, see
Experimental Section).
Figure 4. The two isomers of the cations of the C/Si analogues
2a HCl (2ar/2aβ) and 2b HCl (2br/2bβ).
for protonated sila-trifluperidol (ammonium cation; m/z
426); that is, no disiloxane formation was observed. This
holds true for both freshly prepared samples and solutions
that were kept at 20 °C for 24 h. The stability of the silanol in
3
3
reasonable to assume that 2ar and 2br (with the N-organyl
group in an equatorial position) are the dominating species
in solution. The two isomers (2ar/2aβ; 2br/2bβ) can be
interconverted by a process involving nitrogen deprotona-
tion/reprotonation, nitrogen inversion, and ring inversion.
ESI-MS Studies. To get information about the stability of
2b toward a potential condensation (disiloxane formation) in
aqueous solution, ESI-MS experiments were performed. For
water could also be confirmed for solutions of 2b HCl at
3
higher concentrations (2.5 mM, pH 1 and pH 5; 1 mM, pH
7.4 and pH 10). In any case, no disiloxane formation could be
observed.
Functional Receptor Studies. The inhibitory potencies of
haloperidol (1a), sila-haloperidol (1b), trifluperidol (2a), and
sila-trifluperidol (2b) were studied at human dopamine D1
and D2 receptors using a calcium fluorimetric functional
assay.22 Table 3 presents the Ki values of 1a, 1b, 2a, and 2b
this purpose, 10 μM buffered aqueous solutions of 2b HCl
3
at different pH values (pH 1.0, 5.0, 7.4, and 10.0) were
analyzed. Some results of these studies are depicted in
Figure 5.
The mass spectra of the aqueous solutions of 2b HCl at
3
€
(22) Kassack, M. U.; Hofgen, B.; Lehmann, J.; Eckstein, N.; Quillan,
ꢀ
J. M.; Sadee, W. J. Biomol. Screening 2002, 7, 233–246.
any pH value measured showed only the characteristic peak

1656 Organometallics, Vol. 29, No. 7, 2010
Tacke et al.
Table 3. Inhibitory Potencies of 1a, 1b, 2a, and 2b at Human D1
and D2 Receptors
Table 4. Selectivity of 1a, 1b, 2a, and 2b for Human D1 over
Human D2 Receptors
app. Ki [nM]a
ratio of Ki values
receptor subtype
1a
1b
2a
2b
selectivity
hD1/hD2
1a
1b
2a
2b
hD1
hD2
398
0.36
194
0.12
616
0.024
831
0.24
1110
1620
25 700
3460
a Values result from at least two independent experiments carried out
in triplicate. Hill slopes were not significantly different from unity.
Table 5. Relative Potency of 1b, 2a, and 2b Compared to
Haloperidol (1a)
relative Kia
receptor subtype
1a
1b
2a
2b
hD1
hD2
1.00
1.00
2.05
3.00
0.65
15.0
0.48
1.50
a Values are expressed as ratio of Ki (1a) over Ki (test compound) for
each receptor subtype.
trifluperidol (2a f 2b) reduced this very high D2 selectivity
by around 7-fold, whereas sila-substitution of haloperidol
(1a f 1b) had almost no effect on the D2 selectivity (Table 4).
Table 5 displays the relative potencies of the four com-
pounds studied, with the potency of haloperidol (1a) set as 1.
The remarkably higher potency of 2a compared to 1a found
in this study by using a functional readout is in accordance
with binding studies from the literature.23,24 The reason for
the 10-fold loss of inhibitory potency at D2 receptors upon
sila-substitution of trifluperidol (2a f 2b) remains unknown
and is opposite of what was found for the sila-substitution of
haloperidol (1a f 1b).
Figure 6. Concentration inhibition curves resulting from inhi-
bition of the agonist (SKF38393, 100 nM)-induced signal using
HEK293 cells recombinantly expressing hD1 receptors. Data
are means ( SEM of at least two independent experiments
performed in triplicate. Slopes are not significantly different
from unity.
Conclusions
Sila-trifluperidol (2b), a silicon analogue of the dopamine
(D2) antagonist trifluperidol (2a), was prepared from
triethoxy(vinyl)silane in a multistep synthesis and was iso-
lated as the hydrochloride 2b HCl. In this synthesis, the use
3
of the 2,4,6-trimethoxyphenyl moiety as a protecting group
for silicon played a key role. The silanol 2b was found to be
very stable in aqueous solution. As shown by ESI-MS
studies, no silanol condensation (formation of the corres-
ponding disiloxane) was observed in the pH range 1-10 at
20 °C. Crystal structure analyses of the C/Si analogues
2a HCl and 2b HCl showed that the piperidinium and
Figure 7. Concentration inhibition curves resulting from inhi-
bition of the agonist (quinpirol, 30 nM)-induced signal using
HEK293 cells recombinantly expressing hD2 receptors. Data
are means ( SEM of at least two independent experiments
performed in triplicate. Slopes are not significantly different
from unity.
3
3
4-silapiperidinium ring, respectively, adopt a chair confor-
mation, with the exocyclic N-organyl group and the
3-(trifluoromethyl)phenyl group in equatorial positions.
Sila-substitution of 2a HCl (f 2b HCl) resulted in a more
3
3
“flattened” heterocycle due to the larger covalent radius of
silicon compared to carbon. NMR studies revealed the
existence of two isomers of the cations of 2a HCl and
obtained in these studies. Hill slopes were not significantly
different from unity, thus assuming a single binding site at all
receptors. None of the compounds displayed any agonist
activity at D1 or D2 receptors (not shown). The functional
data for 1a and 1b were similar to the binding data reported
previously.14 Figures 6 and 7 display the inhibition curves of
2a and 2b in comparison to 1a.
Whereas sila-substitution of haloperidol (1a f 1b)
increased the inhibitory potency by 2-fold (D1) and 3-fold
(D2), sila-substitution of trifluperidol (2a f 2b) did not
significantly change the Ki value at D1 receptors but
surprisingly led to a 10-fold reduction of inhibition at D2
receptors.
3
2b HCl in solution (solvent, [D6]DMSO), with molar ratios
3
of 12:1 and 2:1, respectively. These different molar ratios
indicate considerable differences in the energies of the res-
pective two isomers of the piperidinium and silapiperidinium
skeletons. These and other chemical and physiochemical
sila-substitution effects (in this context, see also ref 20) may
be responsible for the striking biological sila-substitution
effects observed for the C/Si analogues trifluperidol (2a) and
(23) Burstein, E. S.; Ma, J.; Wong, S.; Gao, Y.; Pham, E.; Knapp,
A. E.; Nash, N. R.; Olsson, R.; Davis, R. E.; Hacksell, U.; Weiner,
D. M.; Brann, M. R. J. Pharmacol. Exp. Ther. 2005, 315, 1278–1287.
All compounds studied are more than 1000-fold selective
for D2 over D1 receptors (Table 4). However, 2a shows
the highest D2 selectivity (25 700-fold). Sila-substitution of
ꢀ
ꢀ
(24) Fernandez, J.; Alonso, J. M.; Andres, J. I.; Cid, J. M.; Dıaz, A.;
´
Iturrino, L.; Gil, P.; Megens, A.; Sipido, V. K.; Trabanco, A. A. J. Med.
Chem. 2005, 48, 1709–1712.

Article
Organometallics, Vol. 29, No. 7, 2010 1657
sila-trifluperidol (2b). Functional receptor studies with 2a
and 2b at human dopamine D1 and D2 receptors demon-
strated that the carbon/silicon switch led to a 10-fold reduc-
tion of inhibition at D2 receptors, whereas the inhibitory
potency of 2a and 2b at the D1 receptor was not significantly
changed; that is, sila-substitution of 2a (f2b) significantly
reduced the D2 selectivity by a factor of 7. In this context, it is
interesting to note that the sila-substitution of haloperidol 1a
(f1b) resulted in an increased inhibitory potency at both D1
(2-fold) and D2 receptors (3-fold); that is, in this case the
carbon/silicon switch does not affect D2 selectivity. A satis-
factory explanation for this finding is still missing.
In conclusion, the studies presented here again demon-
strate that the carbon/silicon switch strategy is a powerful
tool for drug design. Replacement of a carbon atom by a
silicon atom in a known drug changes the geometric and
electronic properties and thereby the size, shape, conforma-
tional behavior, lipophilicity, and chemical reactivity of the
molecule. This can affect the drug-receptor interaction and
hence change the pharmacodynamics of the drug. Further, as
shown for the C/Si pairs 1a/1b and 2a/2b, carbon/silicon
exchange can also significantly change the phase I and phase
II metabolism of drugs.
(500.1 MHz, [D6]DMSO; data for two isomers): δ 1.07-1.13,
1.28-1.43, and 1.57-1.64 (m, 4 H, SiCH2CH2N), 1.98-2.04
and 2.06-2.12 (m, 2 H, NCH2CH2CH2C), 3.13-3.22 (m, 4 H,
NCH2CH2CH2C), 3.26-3.32, 3.35-3.43, 3.57-3.62, and
3.68-3.70 (m, 4 H, SiCH2CH2N), 6.94 and 6.98 (s, 1 H, OH),
7.34-7.39 (m, 2 H, H-3/H-5, CC6H4F), 7.64-7.69 (m, 1 H, H-5,
SiC6H4(CF3)), 7.79-7.83 (m,
1 H, H-4, SiC6H4(CF3)),
7.96-7.98 (m, 1 H, H-2, SiC6H4(CF3)), 7.99-8.01 (m, 1 H, H-
6, SiC6H4(CF3)), 8.03-8.09 (m, 2 H, H-2/H-6, CC6H4F), 10.6
and 10.7 (br s, 1 H, NH). 13C NMR (125.8 MHz, [D6]DMSO;
data for two isomers, the dominating isomer marked with an
asterisk [*], except for those signals that could not be resolved): δ
9.3 and 11.7* (SiCH2CH2N), 18.0* and 18.3 (NCH2CH2-
CH2C), 35.1 and 35.3* (NCH2CH2CH2C), 50.4 and 51.6*
(SiCH2CH2N), 51.8 and 55.8* (NCH2CH2CH2C), 115.7 (d,
2JCF=21.7 Hz, C-3/C-5, CC6H4F), 124.3 (q, 1JCF=272.7 Hz,
CF3), 126.7-126.9 (m, C-4, SiC6H4(CF3)), 128.6 (q, 2JCF=31.2
Hz, C-3, SiC6H4(CF3)), 128.8* and 128.9 (C-5, SiC6H4(CF3)),
129.7-129.9 (m, C-2, SiC6H4(CF3)), 130.86 and 130.90* (d, 3JCF
4
= 9.7 Hz, C-2/C-6, CC6H4F), 133.2 (d, JCF = 2.7 Hz, C-1,
CC6H4F), 137.0 and 137.4* (C-1, SiC6H4(CF3)), 137.7* and
1
137.8 (C-6, SiC6H4(CF3)), 165.1 (d, JCF = 251.5 Hz, C-4,
CC6H4F), 197.33 and 197.34* (CO). 19F NMR (376.5 MHz,
[D6]DMSO; data for two isomers, the dominating isomer
marked with an asterisk [*]): δ -105.98* and -105.97 (CC6-
H4F), -61.1* and -61.0 (CF3). 29Si NMR (99.4 MHz, [D6]-
DMSO; data for two isomers, the dominating isomer marked
with an asterisk [*]): δ -11.0* and -10.6. Anal. Calcd for
C21H24ClF4NO2Si (461.96): C, 54.60; H, 5.24; N, 3.03. Found:
C, 54.3; H, 5.2; N, 3.0.
Experimental Section
Syntheses. General Procedures. All syntheses were carried
out under dry nitrogen. The organic solvents used were dried
and purified according to standard procedures and stored under
Triethoxy(vinyl)silane (3). This compound was commercially
available.
Preparation of Diethoxy[3-(trifluoromethyl)phenyl]vinylsilane
(4). A solution of [3-(trifluoromethyl)phenyl]magnesium bro-
mide (prepared from 1-bromo-(3-trifluoromethyl)benzene
(50.6 g, 225 mmol) and magnesium turnings (5.67 g, 233 mmol)
in diethyl ether (275 mL)) was added dropwise at 0 °C within
90 min to a stirred solution of 3 (157 g, 825 mmol) in diethyl ether
(400 mL). After the addition was complete, the reaction mixture
was allowed to warm to 20 °C and was then stirred at this
temperature for a further 20 h. n-Pentane (700 mL) was added,
and the resulting suspension was stirred at 20 °C for 1 h. The
precipitate was removed by filtration, washed with n-pentane
(3 ꢀ 50 mL), and discarded. The filtrate and the wash solutions
were combined, the solvents were removed under reduced
pressure, and the residue was distilled under reduced pressure
to give 92.8 g (488 mmol) of the nonreacted starting material 3.
The higher boiling residue was purified by bulb-to-bulb distilla-
tion (bp 75-80 °C/0.1 mbar) to give 4 in 84% yield as a colorless
€
dry nitrogen. A Buchi GKR 50 apparatus was used for the bulb-
to-bulb distillations. Melting points were determined with a
€
Buchi melting point B-540 apparatus using samples in sealed
glass capillaries. The 1H, 13C, 19F, and 29Si NMR spectra were
recorded at 23 °C on a Bruker DRX-300 (1H, 300.1 MHz; 13C,
75.5 MHz; 19F, 282.4 MHz; 29Si, 59.6 MHz), a Bruker Avance
400 (1H, 400.1 MHz; 13C, 100.6 MHz; 19F, 376.5 MHz; 29Si, 79.5
MHz), or a Bruker Avance 500 NMR spectrometer (1H, 500.1
MHz; 13C, 125.8 MHz; 29Si, 99.4 MHz). C6D6, CD2Cl2, or
[D6]DMSO were used as the solvent. Chemical shifts (ppm) were
determined relative to internal C6HD5 (1H, δ 7.28; C6D6), C6D6
(13C, δ 128.0; C6D6), CHDCl2 (1H, δ 5.32; CD2Cl2), CD2Cl2
(13C, δ 53.8; CD2Cl2), [D5]DMSO (1H, δ 2.49; [D6]DMSO),
[D6]DMSO (13C, δ 39.5; [D6]DMSO), external CFCl3 (19F, δ 0;
C6D6, CD2Cl2, [D6]DMSO), or external TMS (29Si, δ 0; C6D6,
CD2Cl2, [D6]DMSO). Analysis and assignment of the 1H NMR
data were supported by 1H,1H gradient-selected COSY, 13C,1H
gradient-selected HMQC and gradient-selected HMBC, and
29Si,1H gradient-selected HMQC (optimized for 2JSiH=7 Hz).
Assignment of the 13C NMR data was supported by DEPT 135
and the aforementioned 13C,1H correlation experiments.
1
liquid (55.0 g, 189 mmol). H NMR (500.1 MHz, CD2Cl2): δ
1.25 (δX), 3.84 (δA), and 3.87 (δB) (ABX3 system, 2JAB=10.3 Hz,
3JAX,BX = 7.0 Hz, 10 H, SiOCHAHBC(HX)3), 5.96 (δA), 6.14
(δB), and 6.22 (δC) (ABC system, 2JAC=4.4 Hz, 3JAB=19.9 Hz,
3JBC=15.0 Hz, 3 H, SiCHBdCHCHA), 7.51-7.55 (m, 1 H, H-5,
Preparation of Trifluperidol Hydrochloride (2a HCl). This
compound was synthesized according to ref 18.
3
SiC6H4(CF3)), 7.68-7.70 (m,
1 H, H-4, SiC6H4(CF3)),
7.84-7.86 (m, 1 H, H-6, SiC6H4(CF3)), 7.90-7.92 (m, 1 H,
Preparation of 4-Hydroxy-1-[4-oxo-4-(4-fluorophenyl)butyl]-
4-[3-(trifluoromethyl)phenyl]-4-silapiperidinium Chloride (2b HCl).
H-2, SiC6H4(CF3)). 13C NMR (125.8 MHz, CD2Cl2): δ 18.5
(OCH2CH3), 59.3 (OCH2CH3), 124.9 (q, JCF = 272.4 Hz,
3
1
A solution of 10 (580 mg, 936 μmol) and 2 M hydrochloric acid
(950 μL, 1.90 mmol of HCl) in acetone (10 mL) was heated under
reflux for 3 h and was then added dropwise at 20 °C with stirring
to diethyl ether (70 mL). The resulting mixture was stirred at
20 °C for a further 10 min, and the precipitate was isolated
by centrifugation and then further purified by resuspension in
diethyl ether (7 mL) and subsequent isolation by centrifugation.
This purification procedure was repeated three times, and the
resulting product was then dried in vacuo (0.1 mbar, 40 °C, 7 h)
3
CF3), 127.1 (q, JCF =3.8 Hz, C-4, SiC6H4(CF3)), 128.5 (C-5,
2
SiC6H4(CF3)), 130.2 (q, JCF = 31.7 Hz, C-3, SiC6H4(CF3)),
131.4 (q, 3JCF=3.8 Hz, C-2, SiC6H4(CF3)), 131.9 (SiCHdCH2),
5
135.5 (C-1, SiC6H4(CF3)), 137.7 (SiCHdCH2), 138.4 (q, JCF
=
1.4 Hz, C-6, SiC6H4(CF3)). 19F NMR (376.5 MHz, CD2Cl2):
δ -63.0. 29Si NMR (99.4 MHz, CD2Cl2): δ -34.6. Anal. Calcd
for C13H17F3O2Si (290.36): C, 53.78; H, 5.90. Found: C, 53.5;
H, 6.0.
to give analytically pure 2b HCl in 74% yield as a colorless solid
Preparation of Ethoxy[3-(trifluoromethyl)phenyl](2,4,6-tri-
methoxyphenyl)vinylsilane (5). A 2.5 M solution of n-butyl-
lithium in hexanes (30 mL, 75 mmol of n-BuLi) was added
dropwise at 20 °C within 10 min to a stirred mixture of 1,3,5-
3
(320 mg, 694 μmol). The product may be recrystallized from
2-propanol/water (15:10 (v/v)) to afford 2b HCl as a colorless
3
crystalline solid in 49% yield; mp 151-152 °C. 1H NMR

1658 Organometallics, Vol. 29, No. 7, 2010
Tacke et al.
trimethoxybenzene (10.2 g, 60.6 mmol), 1,2-bis(dimethylamino)-
ethane (TMEDA; 7.39 g, 63.6 mmol), n-hexane (10 mL), and
diethyl ether (30 mL). The resulting suspension was stirred at
20 °C for 16 h and was then added dropwise at 0 °C within 15 min
to a stirred solution of 4 (17.6 g, 60.6 mmol) in diethyl ether
(60 mL). After the addition was complete, the reaction mixture
was allowed to warm to 20 °C and was stirred at this temperature
for 2.5 h. The precipitate was removed by filtration, washed with
diethyl ether (3 ꢀ 10 mL), and discarded. The filtrate and wash
solutions were combined, the solvents were removed under
reduced pressure, and the oily residue was purified by bulb-
to-bulb distillation (bp 155-160 °C/0.06 mbar) to give 5 in 70%
yield as a colorless viscous liquid (17.5 g, 42.4 mmol). 1H NMR
(500.1 MHz, CD2Cl2): δ 1.21 (δX), 3.76 (δA), and 3.82 (δB)
and 6 (21.0 g, 53.2 mmol) in THF (260 mL) was stirred at 20 °C
for 15 h, followed by the addition of water (55 mL) and a 3 M
aqueous sodium hydroxide solution (100 mL). Subsequently, a
30 wt % aqueous hydrogen peroxide solution (100 mL) was
added dropwise at 20 °C within 15 min to the stirred reaction
mixture, and the resulting mixture was then heated under reflux
for 2 h. Upon cooling to 20 °C, a 0.1 M aqueous solution of
potassium carbonate (500 mL) and dichloromethane (200 mL)
was added. The organic layer was separated, the aqueous layer
was extracted with dichloromethane (3 ꢀ 200 mL), and the
combined organic phases were dried over anhydrous sodium
sulfate. The solvents were removed under reduced pressure, the
byproduct cyclooctane-1,5-diol was separated from the crude
product by bulb-to-bulb distillation (160 °C/0.02 mbar), and the
residue was purified by column chromatography on Al2O3
(neutral, Type 507C, Brockmann I, 150 mesh, 58 A; Aldrich
cat. no. 199974; deactivated with 6 wt % water) using n-hexane/
dichloromethane/ethanol (20:50:4 (v/v/v)) as the eluent. The
relevant fractions were combined, and the solvents were
removed under reduced pressure to give 7 in 67% yield as a
colorless viscous liquid (15.4 g, 35.8 mmol) that solidified after
being left undisturbed at 20 °C for 5 days to give a colorless
crystalline solid; mp 50-51 °C. 1H NMR (500.1 MHz, C6D6): δ
1.71-1.88 (m, 4 H, SiCH2CH2OH), 2.39 (s, 2 H, OH), 3.22 (s, 6
H, o-OCH3, SiC6H2(OCH3)3), 3.44 (s, 3 H, p-OCH3, SiC6H2-
(OCH3)3), 3.88-3.97 (m, 4 H, SiCH2CH2OH), 6.07 (s, 2 H, H-3/
H-5, SiC6H2(OCH3)3), 7.14-7.17 (m, 1 H, H-5, SiC6H4(CF3)),
7.51-7.53 (m, 1 H, H-4, SiC6H4(CF3)), 7.77-7.80 (m, 1 H, H-6,
SiC6H4(CF3)), 8.18-8.19 (m, 1 H, H-2, SiC6H4(CF3)). 13C
NMR (125.8 MHz, C6D6): δ 20.5 (SiCH2CH2OH), 54.6
(o-OCH3, SiC6H2(OCH3)3), 54.7 (p-OCH3, SiC6H2(OCH3)3),
59.8 (SiCH2CH2OH), 91.2 (C-3/C-5, SiC6H2(OCH3)3), 100.7
2
3
(ABX3 system, JAB = 10.2 Hz, JAX,BX = 7.0 Hz, 5 H,
SiOCHAHBC(HX)3), 3.61 (s, 6 H, o-OCH3, SiC6H2(OCH3)3),
3.83 (s, 3 H, p-OCH3, SiC6H2(OCH3)3), 5.76 (δA), 6.08 (δB), and
2
3
6.54 (δC) (ABC system, JAB=4.0 Hz, 3JAC=20.5 Hz, JBC
=
14.7 Hz, 3 H, SiCHCdCHBHA), 6.12 (s, 2 H, H-3/H-5, SiC6H2-
(OCH3)), 7.44-7.47 (m, 1 H, H-5, SiC6H4(CF3)), 7.59-7.61 (m,
1 H, H-4, SiC6H4(CF3)), 7.75-7.78 (m, 1 H, H-6, SiC6H4(CF3)),
7.82-7.83 (m, 1 H, H-2, SiC6H4(CF3)). 13C NMR (125.8 MHz,
CD2Cl2): δ 18.5 (OCH2CH3), 55.4 (o-OCH3, SiC6H2(OCH3)3),
55.6 (p-OCH3, SiC6H2(OCH3)), 59.9 (OCH2CH3), 91.0 (C-3/C-
5, SiC6H2(OCH3)3), 100.7 (C-1, SiC6H2(OCH3)3), 125.2 (q, 1JCF
=272.2 Hz, CF3), 125.6 (q, 3JCF=3.8 Hz, C-4, SiC6H4(CF3)),
2
127.8 (C-5, SiC6H4(CF3)), 129.4 (q, JCF = 31.4 Hz, C-3,
SiC6H4(CF3)), 131.0 (q, JCF = 3.8 Hz, C-2, SiC6H4(CF3)),
3
5
133.2 (SiCHdCH2), 137.4 (SiCHdCH2), 138.0 (q, JCF =1.4
Hz, C-6, SiC6H4(CF3)), 140.1 (C-1, SiC6H4(CF3)), 165.1 (C-4,
SiC6H2(OCH3)3), 167.4 (C-2/C-6, SiC6H2(OCH3)3). 19F NMR
(376.5 MHz, CD2Cl2): δ -62.7. 29Si NMR (99.4 MHz, CD2Cl2):
δ -18.9. Anal. Calcd for C20H23F3O4Si (412.48): C, 58.24; H,
5.62. Found: C, 58.1; H, 5.4.
3
(C-1, SiC6H2(OCH3)3), 125.1 (q, JCF =3.8 Hz, C-4, SiC6H4-
(CF3)), 125.5 (q, JCF =272.4 Hz, CF3), 128.5 (C-5, SiC6H4-
1
2
(CF3)), 129.7 (q, JCF = 31.7 Hz, C-3, SiC6H4(CF3)), 130.6
Preparation of [3-(Trifluoromethyl)phenyl](2,4,6-trimethoxy-
phenyl)divinylsilane (6). A 15 wt % solution of vinylmagnesium
chloride (d=0.97 g/mL; 27.0 mL, 45.3 mmol of CH2dCHMgCl)
in THF was added dropwise at 20 °C within 10 min to a stirred
solution of 5 (16.2 g, 39.3 mmol) in THF (40 mL). After the
addition was complete, the reaction mixture was stirred at 20 °C
for 1 h, followed by sequential addition of a 0.1 M aqueous
solution of potassium carbonate (100 mL) and diethyl ether
(80 mL). The aqueous layer was separated and extracted with
diethyl ether (2 ꢀ 80 mL), and the combined organic layers were
washed with water (1 ꢀ 150 mL) and dried over anhydrous
sodium sulfate. The organic solvents were removed under
reduced pressure, and the oily residue was purified by bulb-
to-bulb distillation (150-155 °C/0.06 mbar) to give 6 in 92%
yield as a colorless viscous liquid (14.2 g, 36.0 mmol). 1H NMR
(500.1 MHz, CD2Cl2): δ 3.60 (s, 6 H, o-OCH3, SiC6H2(OCH3)3),
(q, 3JCF=4.0 Hz, C-2, SiC6H4(CF3)), 137.5 (q, 5JCF=1.4 Hz,
C-6, SiC6H4(CF3)), 141.7 (C-1, SiC6H4(CF3)), 164.7 (C-4,
SiC6H2(OCH3)3), 167.0 (C-2/C-6, SiC6H2(OCH3)3). 19F NMR
(282.4 MHz, C6D6): δ -62.4. 29Si NMR (99.4 MHz, C6D6): δ
-11.3. Anal. Calcd for C20H25F3O5Si (430.50): C, 55.80; H,
5.85. Found: C, 56.2; H, 6.1.
Preparation of 3-[2-(4-Fluorophenyl)-1,3-dioxolan-2-yl]pro-
pylamine (9). This compound was synthesized according to ref 25.
Preparation of Bis[2-(methanesulfonyloxy)ethyl][3-(trifluoro-
methyl)phenyl](2,4,6-trimethoxyphenyl)silane (8) and 4-[3-(Tri-
fluoromethyl)phenyl]-1-{3-[2-(4-fluorophenyl)-1,3-dioxolan-2-yl]-
propyl}-4-(2,4,6-trimethoxyphenyl)-4-silapiperidine (10). Methane-
sulfonyl chloride (1.57 g, 13.7 mmol) was added dropwise
at -25 °C within 10 min to a stirred solution of 7 (2.77 g,
6.43 mmol) and triethylamine (1.89 g, 18.7 mmol) in dichloro-
methane (60 mL). After the addition was complete, the reac-
tion mixture was stirred for a further 10 min at -25 °C, the
mixture was then allowed to warm to 20 °C, and n-pentane
(120 mL) was added. The resulting suspension was stirred at
20 °C for 10 min, and the precipitate was removed by filtration,
washed with n-pentane (3 ꢀ 15 mL), and discarded. The filtrate
and wash solutions were combined, the solvents were removed
under reduced pressure, and the residue was dried in vacuo (0.02
mbar, 30 min) to give 8as a colorless viscous liquid, which was pure
enough to be used in the following step without further purifica-
tion. 1H NMR (500.1 MHz, C6D6): δ 1.74-1.90 (m, 4 H,
SiCH2CH2O), 2.36 (s, 6 H, S(O)2CH3), 3.28 (s, 6 H, o-OCH3,
SiC6H2(OCH3)3), 3.42 (s, 3 H, p-OCH3, SiC6H2(OCH3)3),
4.37-4.49 (m, 4 H, SiCH2CH2O), 6.04 (s, 2 H, H-3/H-5,
SiC6H2(OCH3)3), 7.09-7.12 (m, 1 H, H-5, SiC6H4(CF3)),
7.47-7.49 (m, 1 H, H-4, SiC6H4(CF3)), 7.55-7.57 (m, 1 H,
H-6, SiC6H4(CF3)), 7.99-8.00 (m, 1 H, H-2, SiC6H4(CF3)).
3.83 (s, 3 H, p-OCH3, SiC6H2(OCH3)3), 5.64 (δC), 6.12 (δB), and
=
2
6.61 (δA) (ABC system, JBC=3.8 Hz, 3JAB=14.5 Hz, 3JAC
20.3 Hz, 6 H, SiCHAdCHBHC), 6.12 (s, 2 H, H-3/H-5, SiC6H2-
(OCH3)3), 7.42-7.46 (m, 1 H, H-5, SiC6H4(CF3)), 7.57-7.60
(m, 1 H, H-4, SiC6H4(CF3)), 7.69-7.71 (m, 1 H, H-6, SiC6H4-
(CF3)), 7.76-7.77 (m, 1 H, H-2, SiC6H4(CF3)). 13C NMR (125.8
MHz, CD2Cl2): δ 55.4 (o-OCH3, SiC6H2(OCH3)), 55.6 (p-
OCH3, SiC6H2(OCH3)), 91.2 (C-3/C-5, SiC6H2(OCH3)3),
1
100.5 (C-1, SiC6H2(OCH3)3), 125.1 (q, JCF=272.2 Hz, CF3),
125.4 (q, 3JCF=3.8 Hz, C-4, SiC6H4(CF3)), 127.8 (C-5, SiC6H4-
2
(CF3)), 129.5 (q, JCF=31.2 Hz, C-3, SiC6H4(CF3)), 131.6 (q,
3JCF =3.8 Hz, C-2, SiC6H4(CF3)), 133.6 (SiCHdCH2), 136.6
(SiCHdCH2), 138.7 (q, 5JCF=1.6 Hz, C-6, SiC6H4(CF3)), 139.5
(C-1, SiC6H4(CF3)), 164.9 (C-4, SiC6H2(OCH3)3), 167.1 (C-2/C-
6, SiC6H2(OCH3)3). 19F NMR (376.5 MHz, CD2Cl2): δ -62.7.
29Si NMR (99.4 MHz, CD2Cl2): δ -24.8. Anal. Calcd for
C20H21F3O3Si (394.47): C, 60.90; H, 5.37. Found: C, 60.7; H, 5.4.
Preparation of Bis(2-hydroxyethyl)[3-(trifluoromethyl)phenyl]-
(2,4,6-trimethoxyphenyl)silane (7). A solution of 9-borabicyclo-
[3.3.1]nonane (16.2 g, 66.4 mmol (based on the 9-BBN dimer))
(25) Ismaiel, A. M.; de Los Angeles, J.; Teitler, M.; Ingher, S.;
Glennon, R. A. J. Med. Chem. 1993, 36, 2519–2525.

Article
Organometallics, Vol. 29, No. 7, 2010 1659
13C NMR (125.8 MHz, C6D6): δ 17.9 (SiCH2CH2O), 36.9
(S(O)2CH3), 54.7 (o-OCH3, SiC6H2(OCH3)3), 54.8 (p-OCH3,
SiC6H2(OCH3)3), 68.8 (SiCH2CH2O), 91.0 (C-3/C-5, SiC6H2-
(OCH3)3), 97.4(C-1, SiC6H2(OCH3)3), 125.1 (q, 1JCF=272.4 Hz,
ESI-MS Studies. a. Chemicals. Water (HPLC gradient
grade) was purchased from Acros (Geel, Belgium). Acetic acid
(98%, analytical reagent grade), ammonium hydroxide solution
(25%, analytical reagent grade), and ammonium acetate (ana-
lytical reagent grade) were purchased from Fluka (Taufkirchen,
Germany).
3
CF3), 125.9 (q, JCF =3.8 Hz, C-4, SiC6H4(CF3)), 128.5 (C-5,
2
SiC6H4(CF3)), 130.2 (q, JCF = 32.0 Hz, C-3, SiC6H4(CF3)),
130.3 (q, 3JCF=4.0 Hz, C-2, SiC6H4(CF3)), 137.2 (q, 5JCF=1.2
Hz, C-6, SiC6H4(CF3)), 138.8 (C-1, SiC6H4(CF3)), 165.4 (C-4,
SiC6H2(OCH3)3), 167.0 (C-2/C-6, SiC6H2(OCH3)3). 19F NMR
(376.5 MHz, C6D6): δ -62.1. 29Si NMR (99.4 MHz, C6D6): δ
-13.4. Compound 8 (crude product) was dissolved in aceto-
nitrile (100 mL), triethylamine (2.60 g, 25.7 mmol) and 9 (1.67 g,
7.41 mmol) were added at 20 °C, and the reaction mixture was
then heated at 80 °C for 15 h. The solvent and the excess trie-
thylamine were removed under reduced pressure, and dichloro-
methane (50 mL) and a 0.1 M aqueous solution of potassium
carbonate (50 mL) were added to the residue. The organic layer
was separated, the aqueous layer was extracted with dichloro-
methane (2 ꢀ 50 mL), the combined organic phases were dried
over anhydrous sodium sulfate, and the solvent was removed
under reduced pressure. The residue was purified by column
chromatography on Al2O3 (neutral, Type 507C, Brockmann I,
150 mesh, 58 A; Aldrich cat. no. 199974; deactivated with
6 wt % water) using n-hexane/ethyl acetate/triethylamine
(70:30:1 (v/v/v)) as the eluent. The relevant fractions were
combined and the solvents removed by bulb-to-bulb distillation
(85 °C, 0.02 mbar, 12 h) to give 10 in 37% yield as a colorless,
highly viscous liquid (1.49 g, 2.40 mmol). 1H NMR (500.1 MHz,
C6D6): δ 1.67-1.72 and 1.77-1.90 (m, 6 H, SiCH2CH2N,
NCH2CH2CH2C), 2.18-2.22 (m, 2 H, NCH2CH2CH2C),
b. Sample Preparation. A 10 mM aqueous ammonium ace-
tate buffer was prepared from a 1.0 M stock solution by diluting
with water and adjusting the pH with 98% acetic acid (pH 5),
0.25% ammonium hydroxide solution (pH 7.4), or 2.5% ammo-
nium hydroxide solution (pH 10). For measurements at pH 1,
0.1 M hydrochloric acid was used as the solvent.
Sample solutions of 2b HCl with a concentration of 10 μM
3
were prepared from a 1.0 mM aqueous stock solution by
dilution with 0.1 M hydrochloric acid (pH 1) or with the
respective buffer (pH 5, pH 7.4, pH 10) and were analyzed by
ESI-MS (i) 15 min and (ii) 24 h after preparation.
Sample solutions with concentrations of 2.5 mM (pH 1, pH 5)
and 1 mM (pH 7.4, pH 10) of 2b HCl were prepared by
dissolving the appropriate amount of 2b HCl in 0.1 M hydro-
3
3
chloric acid (pH 1) or in the respective buffer solution (pH 5, pH
7.4, pH 10). At pH 10, acetonitrile was added as a cosolvent
(buffer/acetonitrile, 2:1 (v/v)) for reasons of solubility. The
samples were analyzed by ESI-MS (i) 15 min and (ii) 24 h after
preparation.
c. ESI-MS Analysis. Analysis was performed with a TSQ
7000 tandem mass spectrometer system equipped with an ESI
interface (Finnigan MAT, Bremen, Germany) and a syringe
pump (Harvard Apparatus, South Natick, MA). Data acquisi-
tion and analysis were conducted using Xcalibur Qual Browser
Software 1.2/1.3 (Thermo Electron Corp., Dreieich, Germany).
A flow rate of 10 μL/min was used. The analysis was performed
in the positive ionization mode. The spray capillary voltage was
set to 3.2 kV, and the temperature of the heated capillary was
250 °C. Nitrogen served both as sheath (70 psi) and as auxiliary
gas (10 L/min). The mass spectrometer was operated in the full-
scan mode, m/z 400-1500, with a total scan duration of 1.0 s and
a dwell time of 2 ms.
3
2.44-2.47 (t, JHH = 7.1 Hz, 2 H, NCH2CH2CH2C), 2.75-
2.80 and 3.06-3.11 (m, 4 H, SiCH2CH2N), 3.28 (s, 6 H,
o-OCH3, SiC6H2(OCH3)3), 3.43-3.45 (m, 5 H, p-OCH3, SiC6H2-
(OCH3)3, and CH2OC), 3.66-3.68 (m, 2 H, CH2OC), 6.05 (s,
2 H, H-3/H-5, SiC6H2(OCH3)3), 6.91-6.95 (m, 2 H, H-3/H-5,
CC6H4F), 7.13-7.16 (m, 1 H, H-5, SiC6H4(CF3)), 7.49-7.53
(m, 3 H, H-4, SiC6H4(CF3), and H-2/H-6, CC6H4F), 7.92-7.94
(m, 1 H, H-6, SiC6H4(CF3)), 8.37-8.39 (m, 1 H, H-2, SiC6H4-
(CF3)). 13C NMR(125.8 MHz, C6D6):δ 14.6 (SiCH2CH2N), 22.5
(NCH2CH2CH2C), 39.0 (NCH2CH2CH2C), 53.3 (SiCH2-
CH2N), 54.6 (o-OCH3, SiC6H2(OCH3)3), 54.7 (p-OCH3, SiC6-
H2(OCH3)3), 58.6 (NCH2CH2CH2C), 64.5 (CH2OC), 91.0
(C-3/C-5, SiC6H2(OCH3)3), 102.0 (C-1, SiC6H2(OCH3)3), 110.5
(CH2OC), 115.0 (d, 2JCF=21.2 Hz, C-3/C-5, CC6H4F), 125.4 (q,
Functional Receptor Studies. a. Materials. cDNA for the
hD1 dopamine receptor was kindly provided by Dr. D. Grandy
(Portland, OR). A pcDNA vector construct for hD2S recep-
ꢀ
tors was a gift from Dr. W. Sadee (San Francisco, CA). All
other reagents were purchased from Sigma-Aldrich Chemicals
(Taufkirchen, Germany) unless otherwise stated.
1
3JCF =3.8 Hz, C-4, SiC6H4(CF3)), 125.5 (q, JCF = 272.2 Hz,
b. Cell Cultures. HEK293 cells stably expressing dopamine hD1
or hD2S receptors were established as previously described.26,27
Both cell lines were grown in Dulbecco’s modified Eagle med-
ium (DMEM, GlutaMAX; Invitrogen, Karlsruhe, Germany)
containing 100 U/mL penicillin G, 100 μg/mL streptomycin,
and 10% fetal bovine serum (FBS). Then 400 μg/mL active
G-418 (Calbiochem, San Diego, CA) was added to the medium
of the stably transfected cell lines to maintain selection pressure.
Cells were incubated at 37 °C in a humidified atmosphere under
5% CO2. Confluent cell lines were subcultured by harvesting
with 0.05% trypsin/0.02% EDTA and cultivated in tissue
3
CF3), 128.0 (d, JCF =8.1 Hz, C-2/C-6, CC6H4F), 128.1 (C-5,
SiC6H4(CF3)), 129.9 (q, 2JCF=31.2 Hz, C-3, SiC6H4(CF3)), 131.2
3
(q, JCF = 3.8 Hz, C-2, SiC6H4(CF3)), 138.0-138.1 (m, C-6,
SiC6H4(CF3)), 139.6 (d, 4JCF=3.0 Hz, C-1, CC6H4F), 141.2 (C-1,
1
SiC6H4(CF3)), 162.9 (d, JCF=245.0 Hz, C-4, CC6H4F), 164.5
(C-4, SiC6H2(OCH3)3), 167.0 (C-2/C-6, SiC6H2(OCH3)3). 19F
NMR (376.5 MHz, C6D6): δ -115.0 (CC6H4F), -62.0 (CF3).
29Si NMR (99.4 MHz, C6D6): δ -17.0. Anal. Calcd for
C32H37F4NO5Si (619.73): C, 62.02; H, 6.02; N, 2.26. Found: C,
61.8; H, 5.6; N, 2.4.
2D 1H,1H EXSY NMR Studies. 1H, 13C, 19F, and 29Si NMR
studies of the C/Si analogues 2a HCl and 2b HCl demonstrated
the existence of two isomers of the respective ammonium cations
€
culture flasks (Sarstedt AG & Co., Numbrecht, Germany).
c. Measurement of Changes in Intracellular [Ca2þ] in HEK293
Cells. Measurement of changes in intracellular [Ca2þ] ([Ca2þ]i)
was performed as previously described using a NOVO-
star microplate reader with built-in pipettor (BMG LabTech,
Offenburg, Germany).26 Culture medium was changed approxi-
mately 24 h prior to the assay. A confluent tissue culture flask
(175 cm2) of HEK293 cells expressing the respective dopamine
receptor was loaded with 0.3 μM Oregon Green 488 BAPTA-1/
AM (Molecular Probes, Eugene, OR) for 45 min at 37 °C and
5% CO2 in FBS-free DMEM containing 0.006% (w/v) Pluronic
3
3
(2a HCl, molar ratio ca. 12:1; 2b HCl, molar ratio ca. 2:1) in
3
3
solution (solvent, [D6]DMSO). These isomers are configura-
tionally stable on the NMR time scale under the experimental
conditions used. To prove the existence of the two isomers, 2D
1H,1H EXSY NMR experiments with 2b HCl were carried
3
out at 23 °C (solvent, [D6]DMSO). In this study, the site
exchange for the SiCH2CH2N and SiCH2CH2N protons of
the 4-silapiperidinium skeleton could be observed, showing
strong cross-peaks between the respective signals of the two
isomers. The mixing time was in the order of the spin-lattice
relaxation time T1, calculated by a standard 1D T1-inversion
recovery experiment.
€
(26) Kassack, M. U.; Hofgen, B.; Decker, M.; Eckstein, N.; Lehmann,
J. Naunyn-Schmiedeberg’s Arch. Pharmacol. 2002, 366, 543–550.
(27) Kassack, M. U. AAPS PharmSci. 2002, 4, E31.

1660 Organometallics, Vol. 29, No. 7, 2010
Tacke et al.
F-127. Cells were then harvested with culture medium. After
centrifugation (1200 rpm, 4 °C, 4 min), the cell pellet was rinsed
with two 800 μL portions of Krebs-HEPES buffer (KHB:
118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4,
4.2 mM NaHCO3, 11.7 mM D-glucose, 1.3 mM CaCl2, 10 mM
HEPES, pH 7.4) and then resuspended in 17 mL of the same
solution. The cell suspension was evenly plated into a 96-well
according to the following equation adapted from Cheng and
Prusoff:28
IC50
Ki ¼
L
1 þ
EC50
€
microplate (Sarstedt AG & Co., Numbrecht, Germany). Micro-
where IC50 is the inhibitory concentration 50% of the
antagonist, EC50 is the effective concentration 50% of the
agonist used (SKF38393 for hD1 and quinpirole for hD2
receptors), and L is the molar concentration of the agonist
used. Data are presented as means ( SD (n g 2).
plates were kept at 37 °C for 30 min before antagonists,
dissolved in KHB containing 0.5% bovine serum albumine
(BSA), were added. After a further 30 min incubation at 37 °C,
fluorescence intensity was measured at 520 nm (bandwidth
35 nm) for 5 s at 1.0 s intervals to monitor baseline. Control
solutions or agonist solutions (hD1: 100 nM SKF38393; hD2:
30 nM quinpirole), dissolved in KHB containing 0.5% BSA,
were then injected into separate wells, and fluorescence inten-
sity was monitored at 520 nm (bandwidth 35 nm) for 20 s
at 0.4 s intervals. Excitation wavelength was 485 nm (band-
width 12 nm).
Acknowledgment. Experimental support for the ESI-
€
MS studies by W. Hummer, Lehrstuhl Lebensmittel-
€
chemie, Universitat Wurzburg, is gratefully acknowl-
edged. C.U. was supported by the Dr. Hilmer foundation.
€
d. Data Analysis. The change in [Ca2þ]i is expressed as a
change in fluorescence intensity. Data from at least two inde-
pendent experiments each assayed in triplicates were pooled
after normalization, and concentration-inhibition curves were
constructed using nonlinear regression curve fit (sigmoidal
dose-response equation) by means of Prism software 4.0
from GraphPad (La Jolla, CA). Ki values were calculated
Supporting Information Available: Crystallographic data for
2a HCl and 2b HCl and inhibitory potencies (pKi values (
3
3
SEM) of 1a, 1b, 2a, and 2b. This material is available free of
charge via the Internet at http://pubs.acs.org.
(28) Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099–
3108.