A new sif-dipropargyl glycerol scaffold as a versatile prosthetic group to design dimeric radioligands: Synthesis of the [18f]bmppsif tracer to image serotonin receptors

Article abstract of DOI:10.1002/cmdc.201300458

A novel SiX-dipropargyl glycerol scaffold (X: H, F, or 18F) was developed as a versatile prosthetic group that provides technical advantages for the preparation of dimeric radioligands based on silicon fluoride acceptor pre- or post-labeling with fluorine

Full text of DOI:10.1002/cmdc.201300458

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201300458
A New SiF–Dipropargyl Glycerol Scaffold as a Versatile
Prosthetic Group to Design Dimeric Radioligands:
Synthesis of the [18F]BMPPSiF Tracer to Image Serotonin
Receptors
Puja Panwar Hazari,[a] Jurgen Schulz,[b] Delphine Vimont,[b] Nidhi Chadha,[a] Michele Allard,[b]
Magali Szlosek-Pinaud,[c] Eric Fouquet,[c] and Anil Kumar Mishra*[a]
A novel SiX–dipropargyl glycerol scaffold (X: H, F, or 18F) was
developed as a versatile prosthetic group that provides techni-
cal advantages for the preparation of dimeric radioligands
based on silicon fluoride acceptor pre- or post-labeling with
fluorine-18. Rapid conjugation with the prosthetic group takes
place in microwave-assisted click conjugation under mild con-
ditions. Thus, a bivalent homodimeric SiX–dipropargyl glycerol
derivatized radioligand, [18F]BMPPSiF, with enhanced affinity
was developed by using click conjugation. High uptake of the
radioligand was demonstrated in 5-HT1A receptor-rich regions
in the brain with positron emission tomography. Molecular
docking studies (rigid protein–flexible ligand) of BMPPSiF and
known antagonists (WAY-100635, MPPF, and MefWAY) with
monomeric, dimeric, and multimeric 5-HT1A receptor models
were performed, with the highest G score obtained for docked
BMPPSiF: À6.766 as compared with all three antagonists on
the monomeric model. Multimeric induced-fit docking was
also performed to visualize the comparable mode of binding
under in vivo conditions, and a notably improved G score of
À8.455 was observed for BMPPSiF. These data directly correlate
the high binding potential of BMPPSiF with the bivalent bind-
ing mode obtained in the biological studies. The present study
warrants wide application of the SiX–dipropargyl glycerol pros-
thetic group in the development of ligands for imaging with
enhanced affinity markers for specific targeting based on pep-
tides, nucleosides, and lipids.
Introduction
Positron emission tomography (PET) is an appropriate tool to
measure and quantify the uptake and kinetics of radiotracers
for neuroreceptors as well as to evaluate the efficacy of drugs
directed toward these molecular targets. Introduction of 18F
into biomacromolecules (peptides, aptamers, nucleosides, etc.)
ration of synthesis.[1–5] CuACC redundant-site-specific methods
for the 18F labeling of peptides have also been reported and in-
cluded a) the use of 18F-fluorinated aldehydes with hydrazine-
modified peptides,[6] 18F-active esters, and other reactive
groups[7] and b) the chemoselective modification of thiol
groups in peptides with maleimide-functionalized prosthetic
18
18
is a challenge with CÀ F chemistry. CÀ F chemistry has the
limitation of long and complicated synthetic procedures, and
conjugation to proteins and purification requires more than
one half-life of the 18F radioisotope. Peptide labeling by using
an efficient click reaction with a CuI-catalyzed azide–alkyne 1,3-
18
groups, which provides CÀ F bond formation.[8] Other novel
strategies to facilitate the introduction of 18F labels into pep-
tides and larger molecules have been well demonstrated by
Schirrmacher and co-workers.[9] The discovery of reliable meth-
ods for 18F labeling that can be applied by technicians on
a daily basis would be a critical step toward an improvement
in producing PET radiopharmaceuticals for varied applications.
One-step labeling with a prosthetic group is of great inter-
est, primarily for the 18F labeling of proteins and peptides be-
cause they allow radiosyntheses in aqueous systems. As
a result, 18F-labeled prosthetic groups were successfully cou-
pled with larger biomolecules. One such method involves 18F
incorporation in a single step due to silicon-based fluoride ac-
ceptor (SiFA) modified bioactive probes[10,11] and through an
aluminum conjugate (Al[18F]),[12,13] which has potential advan-
18
dipolar cycloaddition (CuACC) that relies on CÀ F bond forma-
tion through 18F-fluorinated azide- or alkyne-containing build-
ing blocks has been reported and provides the technical ad-
vantages of unparalleled specific activity and a reasonable du-
[a] Dr. P. P. Hazari, N. Chadha, Dr. A. K. Mishra
Division of Cyclotron and Radiopharamceutical Sciences
Brig. SK Mazumdar Road, Delhi 110054 (India)
E-mail: akmishra63@gmail.com
[b] Dr. J. Schulz, D. Vimont, Prof. M. Allard
CNRS, INCIA, UMR 5287, 33400 Talence (France)
18
[c] Dr. M. Szlosek-Pinaud, Prof. E. Fouquet
University of Bordeaux, ISM, UMR 5255
Synthesis—Bioactive Molecules Group 351
Cours de la liberation, 33405 Talence (France)
tages over CÀ F radiochemistry.
Direct 18F-labeling approaches for complex biomolecules are
difficult to achieve, so the most recent SiFA labeling protocols
for peptide labeling are easy and convenient, even for a pros-
thetic group for direct radiolabeling of bioactive conjugates
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201300458.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 337

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
for application in positron emission tomography.[14–16] More-
over, unique and efficient one-step approaches based on sili-
con chemistry for the 18F fluorination of nucleosides and the
RGD peptide, respectively, have been demonstrated previously;
these methods lead to high radiochemical purity accompanied
with a high product purity and represent a very promising al-
ternative to the actual 18F-labeled radiotracers.[17,18]
ligand through click chemistry and then its 18F labeling by SiFA
technology and imaging of the serotonin receptor by PET.
Results and Discussion
The present study provides a versatile prosthetic group, SiF–di-
propargyl glycerol, that is usable in the Huisgen reaction,
which enables 18F labeling with a small quantity of azide com-
pound made available as a counterpart substrate. The stepwise
synthesis of the SiF conjugated building block involved the re-
action of di-tert-butylchlorosilane with 4-bromophenol to yield
4-(di-tert-butylsilanyl)phenol (1) by using a procedure, with
modification,[17,18] already described by Schreiber and co-work-
ers.[26,27] The silylated phenol was then treated with solketal to
introduce the glycerol scaffold in 63% yield. The resultant diol
was obtained by a deprotection step and treated with prop-
argyl bromide to render bivalent functionality for further con-
jugation. Thus, (4-(2,3-bis(prop-2-ynyloxy)propoxy)phenyl)-di-
tert-butylsilane (SiH–dipropargyl glycerol, 4) was obtained in
81% chemical yield (Scheme 1). The second stage involves the
A variety of neurodiagnostic PET tracers have already been
reported in literature. But these have inadequacies, like insuffi-
cient blood–brain-barrier passage and susceptibility to meta-
bolic cleavage. A highly attractive approach to investigate and
control GPCR dimerization may be provided by the exploration
and characterization of bivalent ligands, which can act as mo-
lecular probes simultaneously binding two adjacent binding
sites of a dimer.[19,20] Monomeric 5-HT1A receptor PET ligands
with better metabolic profiles and longer radioactive half-lives
have been reported in the literature, namely the fluoro and
bromo analogues of WAY-100635 (N-[2-[4-(2-methoxyphenyl)-1-
piperazinyl]ethyl]-N-(2-pyridinyl)cyclohexanecarboxamide), that
is, 6FPWAY (N-(2-(1-(4-(2-methoxyphenyl)piperazinyl)ethyl))-N-
(2-(6-fluoropyridinyl))cyclohexanecarboxamide) and 6BPWAY
(N-(2-(1-(4-(2-methoxyphenyl)piperazinyl)ethyl))-N-(2-(6-bromo-
pyridinyl))cyclohexanecarboxamide).[21–23] [18F]6FPWAY demon-
strated resistance to defluorination, but it is worth noting that
only moderately high uptake in 5-HT1A receptor rich regions
was achieved. Therefore, this radioligand has not seen use in
human subjects.[22] [18F]MPPF (2’-methoxyphenyl-(N-2’-pyridin-
yl)-p-[18F]fluorobenzamidoethylpiperazine) is another successful
5-HT1A PET ligand, which acts as a reversible, competitive an-
tagonist at 5HT1A receptors and exhibits receptor affinity (dis-
sociation constant Kd =50.3 nm). In an attempt to produce an
18F-labeled ligand that would be stable to defluorination in
vivo, [18F]Mefway (4-([18F]fluoromethyl)-N-{2-[4-(2-methoxy-
phenyl) piperazin-1-yl]ethyl}-N-(pyridin-2-yl)cyclohexane-1-car-
boxamide) was developed. It possesses very similar affinity and
gives a higher target to background radioactivity ratio than
many other 5-HT1A radioligands, including [18F]MPPF. In nonhu-
man primates, the signal is similar to that obtained with [car-
bonyl-11C]WAY-100635 and superior to that of [18F]MPPF.[24]
[11C]WAY-100635 is the current “gold standard” and difficult to
Scheme 1. Synthesis of (4-(2,3-bis(prop-2-ynyloxy)propoxy)phenyl)di-tert-bu-
tylsilane (4). Reagents and conditions: a) PPh3, DEAD, THF, reflux, 48 h, 63%;
b) Dowex 50WX8-200, MeOH, RT, 48 h, 96%; c) NaH, propargyl bromide, THF,
reflux, 15 h, 81%. DEAD: diethylazodicarboxylate.
alkylation of 2-methoxyphenylpiperazine with bromoethanol
and subsequent azide formation (Scheme 2). The optimized
coupling between the organosilicon-functionalized alkyne and
the azide by click chemistry was performed at 1008C under mi-
crowaves (50 W) for 15 min (Scheme 3). Microwave-assisted
synthesize, so
a suitable fluorine-18 replacement, as in
[18F]Mefway, is highly desirable with one-step radiolabeling.
Herein, we have tried to introduce an indirect method of 18F
fluorination that proceeds through a prosthetic group into
a single-step labeling method by isotope/isosteric exchange.
The work puts forth a new route for the development of PET
pharmaceuticals based on the introduction of a prosthetic
group that can be used in direct labeling with 18F. The ap-
proach envisaged is based on the attachment of a suitable
azide functionality linked to a homodimeric bivalent 5-HT1A re-
ceptor ligand that reacts with the terminal SiH–dipropargyl
glycerol prosthetic group in a site-selective manner within
a timeframe acceptable for PET-radiopharmaceutical produc-
tion. We have already demonstrated “click chemistry” involving
[11C]methyl azide, as a more practical, easier, and general alter-
native labeling strategy for labeling biomolecules.[25] Herein we
present the direct coupling of a homobivalent 5-HT1A receptor
Scheme 2. Synthesis of 1-(2-azidoethyl)-4-(2-methoxyphenyl)piperazine (2’).
Reagents and conditions: a) K2CO3, MeCN, 2-bromoethanol, reflux, 48 h, 97%;
b) DPPA, DBU, DMF, 658C, overnight, 91%. DPPA: diphenylphosphorylazide;
DBU: diazabicycloundecene; DMF: N,N-dimethylformamide.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 338

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Scheme 3. Synthesis of 4,4’-(2,2’-(4,4’-(3-(4-(di-tert-butylsilyl)phenoxy)propane-1,2-diyl)bis(oxy)bis(methylene)bis(1H-1,2,3-triazole-4,1-diyl))bis(ethane-2,1-diyl))-
bis(1-(2-methoxyphenyl)piperazine). Reagents and conditions: a) CuSO4, ascorbic acid, tBuOH/H2O (3:1), MW, 10 min, 1008C; b) KF, Kryptofix[2.2.2], AcOH,
DMSO, 15 h, 808C; c) 18FÀ, Kryptofix[2.2.2], AcOH, DMSO, 15 min, 808C.
alkyne–azide cycloaddition provides a possible means of over-
coming some of the difficulties with aqueous solvents that
have been observed with the traditional methodology. The
final labeling precursor, 5, bearing the SiH group was obtained
in 70% yield. Finally, cold fluorination of 5 realizes the forma-
tion of 6 and demonstrates isotope-element labeling with 19F.
The presence of the silyl group in the labeling precursor 5 was
confirmed by the peak at d=12.50 ppm in the proton-decou-
pled 29Si NMR spectrum (Figure 1a).
of the cells had survived after 2 h of treatment with 5, and the
cells were found to regain proliferation at 24 h with a surviving
fraction of 0.896Æ0.020. A surviving fraction of 0.56Æ0.035
was seen at 1 mm concentration for compound 5 (Figure 2).
Dose-dependent and time-dependent growth inhibition of
brain cells, with IC50 values ranging between 1 and 10 mm, was
observed for both of the compounds by clonogenic assay. A
direct correlation was seen between the MTT and clonogenic
The fluorinated derivative 6 was subjected to proton-decou-
pled 29Si and 19F NMR spectroscopy. The 29Si NMR spectrum re-
vealed splitting of the signal due the presence of the fluorine
1
atom at d=14.37 ppm (d, JSiÀF =296.5 Hz). The 19F NMR spec-
trum depicted a chemical shift at d=189.60 ppm and two
1
small satellite peaks with JSiÀF =297 Hz due to the silyl cou-
pling (Figure 1b and c).
We next analyzed the toxicity of compounds 5 and 6 under
in vitro conditions for their cytotoxicity on brain cell homoge-
nate. A similar pattern of survival was observed for both deriv-
atives on brain homogenate. At 0.1 mm concentration, 31.4%
Figure 2. Cytotoxic effect of 5 and 6 in rat brain homogenates. The standard
deviation was always <10%.
assays. The above observations exhibit the useful imaging ap-
plications of these compounds by using nuclear medicine
techniques. For the above synthesized and purified compound,
further implementation into a fully automated system would
be interesting for evaluation of its feasibility under biological
systems.
The optimum protocol for radiolabeling of 5 and 6 consisted
of two parts. First, [18F]À ions were extracted and entrapped in
an anion-exchange resin (light quaternary methylamine car-
tridge) preconditioned with 0.5mK2CO3 and water and eluted
with a complex of Kryptofix[2.2.2] and K2CO3. After removal of
the solvents, the precursors 5 or 6 (6 mg) and glacial acetic
acid (3 mL) in anhydrous DMSO (300 mL) were treated with
dried Kryptofix[2.2.2]–K[18F] at 808C for 15 min. The formation
Figure 1. a) Proton-decoupled 29Si NMR spectrum of 5; b) proton-decoupled
29Si NMR spectrum of 6; c) 19F NMR spectrum of 6.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 339

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
of [18F]BMPPSiF ([18F]5 or [18F]6) was confirmed by radio-HPLC
analysis on a semipreparative column at a flow rate of
2 mLminÀ1. The chemical and radiochemical purity (RCP) were
checked by reinjection of the product onto a radio-HPLC
column; the radioactive fraction was analyzed, collected, and
measured. The [18F]BMPPSiF was well separated from its label-
ing precursor 5 (Figure 3a). The retention time of [18F]BMPPSiF
in good yield. The design of the homodimeric system has
been made on the basis of structure–activity relationship stud-
ies reported in the literature for enhanced affinities toward
GPCR receptors.
To substantiate a receptor-chelating binding mode for
[18F]BMPPSiF, the Hill coefficient (nH) was calculated. The com-
petition curve revealed steepening and led to a Hill slope (nH)
of 2.2, indicative of bivalent binding in which there is a release
of two equivalents of radioligand. The steepening of the com-
petition curve is suggestive of a positive cooperative binding
(Figure 4a). Bivalent ligands addressing two adjacent binding
sites of receptor dimers will induce such cooperativity because
binding of the second pharmacophore is significantly acceler-
ated due to the vicinity of the ligand and the facilitated enrich-
ment of local concentration. This has been demonstrated earli-
er by our group for the bivalent homodimeric 5-HT1A ligand.[28]
Kꢁhhorn et al. synthesized bivalent dopamine D2 receptor li-
gands by incorporating the privileged structure of 1,4-disubsti-
tuted aromatic piperidines/piperazines (1,4-DAPs) and triazolyl-
linked spacer elements. They have also reported radioligand
binding studies and performed a comparative analysis with the
respective monomers and unsymmetrically substituted ana-
logues, which indicated a bivalent binding mode with simulta-
neous occupancy of two adjacent binding sites.[29] It is as-
Figure 3. a) HPLC chromatograms of radiolabeled BMPPSiF after radiosynthe-
sis; b) HPLC chromatograms of [18F]BMPPSiF after HPLC purification.
was 6.8 min (Figure 3b). After purification and formulation,
[18F]BMPPSiF was obtained in (52Æ10.5)% (decay corrected,
n=5) radiochemical yield, with a total synthesis time of (58Æ
3.4) min; the specific activity was (480.6Æ8.6) GBqmmolÀ1 at
the end of the synthesis, and a radiochemical purity of 99%
was obtained. A catalytic amount of acetic acid was suitable
for the 18F fluorination to achieve a decay-corrected synthetic
yield of >55%. Radiolabeling with 6 by isotope exchange pro-
vided 5% more yield at low temperatures. The purpose of la-
beling 6 was to fully demonstrate the capability of the 18F fluo-
rination and to provide flexibility in the system for radiolabel-
ing peptides and larger molecules for application in PET. Nota-
bly, this methodology surpasses the long and complicated con-
18
jugation procedures and provides the SiÀ F bond formation
Figure 4. a) Representative binding curve of [18F]BMPPSiF on the primary
hippocampal culture. This results in an absolute value of the Hill slope (nH)
of 2.2, which is indicative of a bivalent binding mode. b) Kd determination
by using a Scatchard plot of the specific binding data to determine the ratio
of bound to free [18F]BMPPSiF.
during a short synthesis time of 5–15 min with labeled
BMPPSiF of high specific activity.
We present an efficient strategy for the reliable and efficient
synthesis of a bivalent ligand of 1-(2-methoxyphenyl)piperazine
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 340

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
sumed that duplication of the pharmacophoric groups accord-
ing to the bivalent ligand approach leads to a supra-additive
increase in potency relative to that of the corresponding mon-
ovalent ligand.
A sub-picomolar affinity of (59.34Æ2.3) pm for 5-HT1A was
obtained in an in vitro binding assay on primary hippocampal
cultures (Figure 4b); this value was sufficiently high to carry
out imaging studies. The selectivity of the complex for 5-HT1A
and 5-HTT receptors is more than 1000-fold as high as it is for
5-HT2A receptors. The 5-HT2A and D2 receptor affinities of
[18F]BMPPSiF were found to be lower than those observed for
5-HT1A receptors in brain homogenates.
In an in vitro binding assay on rat brain homogenate (for 5-
HT1A, 5-HT2A, D2, and 5-HTT), a sub-nanomolar affinity for 5-
HT1A was obtained. The affinity for the 5-HT2A receptor was ap-
proximately 1000-fold lower in magnitude. The selectivity of
the complex for 5-HT1A receptors showed high affinity with
a value of (0.099Æ0.002) nm, and the affinity toward the sero-
tonin transporter (5-HTT) was found to be approximately 100-
fold lower, which suggested selectivity toward the target
(Table 1).
Table 1. Binding affinity of [18F]BMPPSiF for the 5-HT1A, 5-HT2A, D2, and 5-
HTT receptors.[a]
Receptor
Kd [nm]
Figure 5. Integral 3D PET images of a Sprague–Dawley rat brain co-regis-
tered with the CT scan in the a) anterior and b) posterior. Planar images of
a rat brain in the c) sagittal, d) horizontal, and e) transverse planes. These are
0–60 min summation images after the intravenous injection of [18F]BMPPSiF
and show strong accumulation of radioactivity in the hippocampus (HIP),
cingulate cortex (CG), and caudate putamen (CPu) and moderate accumula-
tion in the nucleus raphe dorsalis.
5-HT1A
5-HT2A
D2
0.099Æ0.02
176Æ8.6
2.2Æ0.98
9.8Æ1.23
5-HTT
[a] Nonspecific binding was determined by using serotonin, ketanserin,
spiperone, and paroxetin for the 5-HT1A, 5-HT2A, D2, and 5-HTT receptors,
respectively.
sults were obtained, which directly correlated with the bivalent
mode of binding to the 5-HT1A receptor in hippocampal cul-
tures. There was a significant decrease in the binding potential
of 1.96 for [18F]MPPF, relative to that of [18F]BMPPSiF.
After injection of [18F]BMPPSiF into Sprague–Dawley (SD)
rats, 9% of the total radioactivity injected accumulated linearly
in the brain over 10 min. The radioactivity distribution was in
concurrence with the known 5-HT1A receptor distribution, with
high uptake of radioactivity in the hippocampus (HIP), cingu-
late cortex (CG), and caudate putamen (CPu) and moderate ac-
cumulation in the nucleus raphe dorsalis (Figure 5). The corre-
sponding binding ratios were 23.0, 16.9, and 12.3 at the time
of peak equilibrium, which was reached approximately 15 min
after injection. A time–activity curve (TAC) of the hippocampus
region depicted maximum accumulation at 15 min and decline
of radioactivity after 30 min. A pretreatment experiment with
serotonin decreased the level of radioactivity in all examined
regions. A region of interest (ROI) analysis of the TAC at 60 min
by means of a simplified reference tissue described elsewhere
yielded a 5-HT1A binding potential (BP) value of 5.99 in the hip-
pocampal region, calculated as BP=Vd(HIP)/Vd(cerebellum), in
which Vd is the volume of distribution. The cerebellum was
used as the reference region for the calculation of BP due to
the low 5-HT1A receptor density in that region. A direct com-
parison with monomeric [18F]MPPF was made due to unavaila-
bility of a bivalent PET radioligand for 5-HT1A. Astonishing re-
Biodistribution studies in SD rats showed that the initial
uptake in the brain was high (2% dose per g of tissue at
2 min). Rates of elimination were significantly slower in brain
regions with high concentrations of 5-HT1A receptors (hippo-
campus, cortex, and hypothalamus) than in control regions,
from which the radioligand was eliminated at a fast pace. Radi-
oactivity in the bone did not increase with time, which sug-
gests that in vivo defluorination may not be the major route of
metabolism. Metabolic degradation is expected for all mole-
cules, but the triazoles in the system provide high aromatic
stabilization and are stable toward acidic and basic hydrolysis.
Triazoles are also resistant to metabolic degradation, as well as
to reducing and oxidizing conditions. Plasma analysis showed
no major degradation of the [18F]BMPPSiF on radio-TLC after
1 h.
Regional brain distribution of [18F]BMPPSiF in whole-brain
sections of SD rats showed rapid accumulation within 10 min,
and a value of 3.92% of the total injected dose per g of tissue
was attained (Figure 6). High uptake in the hippocampus was
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 341

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
For docking studies for the 5-HT monomer model, a known
5-HT1A antagonist binding site was explored for BMPPSiF. Three
antagonists, MefWAY, WAY-100635, and MPPF (Figure 7), were
used in the docking studies to understand the monomer bind-
ing pose of BMPPSiF. These ligands are used in imaging studies
for the 5-HT1A receptor; hence, we used the 18F analogues of
MPPF and MefWAY and considered WAY-100635 to be 11C la-
beled. BMPPSiF was also used in the 18F-labeled form during
the docking studies.
Figure 6. Regional distribution of [18F]BMPPSiF in SD rat brains at 10, 30, and
60 min post-injection. WBS: whole-brain section; OB: olfactory bulb; CPu:
caudate putamen; Hip: hippocampus; Th: thalamus; Ctx: frontal and parietal
cortex; Cb: cerebellum. Data are the mean Æstandard deviation of n=5 ex-
periments.
accredited to the 5-HT1A receptor rich regions. Moreover, in se-
rotonin-depleted rat models,[30] there was significant decrease
in the uptake for [18F]BMPPSiF in the hippocampus region
((1.58Æ0.56)% was observed at 10 min) and a significant in-
crease of (2.98Æ0.16)% at the cortex region at 10 min. In
a blocking experiment, there was a slight decrease in the
uptake for the hippocampus and cortex, which was statistically
insignificant (P>0.05). The high specific affinity and bivalent
mode of binding to 5-HT1A receptors in the brain could be the
main reason for the insignificant change in the blocking re-
sults. There were no significant changes in the activity in other
major organs with the blocking agent.
Figure 7. Structures of antagonists for the serotonin 1A receptor. The antag-
onists MefWAY, WAY-100635, and MPPF were used in the docking studies.
Glide extra precision (XP) mode docking poses were
screened out, based on the minimized energy complex in
terms of the G score. From the docking pose, the antagonist
binding pocket was found to be relatively shallow and in-
volved in various interactions at the site. The G score signifies
better binding of these ligands at the site, and the highest
score was obtained for docked BMPPSiF, with a value of
À6.766, relative to all three antagonists. Docking of ligands in
the cavity involves important hydrophobic/nonpolar residues,
such as Val117, Cys120, Ile189, Thr196, Thr200, Ala203, Phe361,
Phe362, Val364, Ala365, Leu368, Thr188, Tyr195, and Gly382,
and the polar residues Lys191, Ser199, and Ser190. The antago-
nist MPPF is involved in p–p interactions with Phe362 and hy-
drogen bonding with the backbone of Ile189, whereas
MefWAY is involved in p–p interactions with Arg181 and hy-
drogen bonding with the backbone of Ile189. But, in case of
WAY-100635, additional p–p interactions with Phe361 were ob-
tained. BMPPSiF was aligned in the hydrophobic pocket with
one of the bivalent moieties perfectly inserted into the hydro-
phobic environment of the above-described residues. Overall,
the monomer 5-HT1A docked poses depicted better binding
energy for BMPPSiF relative to those of radiolabeled analogues
of known antagonists.
Homology modeling and molecular docking studies with
monomeric, homodimeric, and multimeric 5-HT1A receptor
models
To gain insight into the binding poses and binding affinity of
the bivalent ligand for the serotonin 5-HT1A receptor, homolo-
gy modeling assisted docking studies were performed with
monomeric, homodimeric, and multimeric 5-HT1A receptor
models. We built a reliable monomer model based on the crys-
tal structure of the b2-adrenergic receptor, 2RH1, after a fair se-
quence alignment between the two receptors with approxi-
mately 48% amino acid similarity in the conserved transmem-
brane (TM) region of the template and the 5-HT1A sequence.
The crystal structure of 2RH1 has been the PDB of choice for
modeling of aminergic GPCRs with highest resolution of 2.4.[31]
Reliable modeling of the TM region is important because the
ligand binding pocket lies in the TM region of the 5-HT1A mo-
nomer. The 5-HT1A monomer model was minimized; loops
were refined and evaluated by Ramachandran Plot and PRO-
CHECK server. The model was found to be validated by having
more than 94% of residues fall under allowed regions without
any steric clashes.
The monomer 5-HT1A receptor binding results prompted us
to explore the binding pose and energy of the bivalent ligand
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 342

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 8. Ramachandran Plots: a) dimeric model; b) multimeric model.
BMPPSiF with dimeric and multimeric domains of 5-HT1A recep-
tor models. From experimental studies, the dimerization of
GPCRs is mediated by interactions between TM regions, in
either an intradimeric or interdimeric way. Hence, from the mo-
nomer 5-HT1A model, a dimeric model was built by superim-
posing TM4 and TM5 and validated by Ramachandran Plot and
PROCHECK server (Figure 8).
for all ligands. Site IV was involved in chain E of the model,
with the highest site score of 0.991 and significant hydropho-
bic and hydrophilic contents. Binding of BMPPSiF and other
antagonists on the multimeric site is depicted in Figure 10 (see
also the Supporting Information). At this multimeric site IV,
BMPPSiF was found to show the highest G score of À7.727, as
compared with those of the other known antagonists for 5-
HT1A (Table 2).
The binding sites of homodimers of 5-HT1A were predicted
by using Sitemap. Five potential sites comprising different as-
pects of hydrophobic, hydrophilic, H-bond acceptors, and H-
bond donors were obtained. Docking studies were performed
on grids of all five sites. Out of the five sites, site III was
screened out for the best-docked site on the basis of the high
docking score obtained for all ligands. Site III was involved in
chain B of the model, with the highest site score of 0.9943 and
significant hydrophobic and hydrophilic contents. In dimeric si-
te III, BMPPSiF was found to show the highest G score of
À6.341, as compared with those of other molecules.
Table 2. Predicted binding energies of docked poses of antagonists to
the human serotonin 1A receptor.
Ligand
XP G score
dimeric
monomeric
multimeric
BMPPSiF
MPPF
MefWAY
WAY-100635
À6.766
À5.452
À5.775
À5.829
À6.341
À5.613
À5.630
À4.726
À7.727
À4.517
À5.107
À3.199
The docked BMPPSiF cavity involves important hydrophobic/
nonpolar residues, such as Phe246, Ile239, Met272, and
Met293, and the polar residues Asn243, His229, and Gln271
(Figure 9). Hydrogen bonding was present between the NH
group of the piperazine ring and Asp232. This was found to be
an additional interaction in BMPPSiF, which was lacking in the
other three molecules.
Site IV gave the highest G score for BMPPSiF among the
monomeric, dimeric, and multimeric models due to the good
binding pose of BMPPSiF in this site relative to other binding
poses. The docking pose revealed complete enclosure of
BMPPSiF at site IV and, hence, the docked complex is stabilized
more than the other ligands (Figure 10). The important interac-
tions involved are p–p interactions with Asp339 and hydrogen
bonding with Glu340. Hence, complete enclosure and more in-
teractions at the binding site of BMPPSiF resulted in a high
binding score.
In similar terms, a multimeric 5-HT1A model was prepared
and validated through Ramachandran Plot and PROCHECK
server (Figure 8). Sitemap was used to obtain the binding sites
of the multimeric 5-HT1A model. Five potential sites, comprising
different aspects of hydrophobic, hydrophilic, H-bond accept-
ors, and H-bond donors, were obtained and used for docking
analysis.
Multimeric induced-fit docking was also performed to mimic
the comparable mode of binding under in vivo conditions, and
interestingly, an improved G score of À8.455 was observed for
BMPPSiF. The G score values for the studied 5-HT1A antagonists
Out of all five sites, site IV was screened out for the best
docked site on the basis of the high docking score obtained
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 343

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 9. 3D interactions of ligands (gray surfaces) and homodimeric 5-HT1A receptors (blue surfaces).
Figure 10. 3D interactions of ligands (gray surfaces) and multimeric 5-HT1A receptors (blue surfaces).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 344

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Semipreparative HPLC purifications were carried out by using
a semipreparative Phenomenex Luna C18 column (250ꢂ10 mm,
5 mm, 2.5 mLminÀ1) under the indicated conditions. The semipre-
parative HPLC system used a Sykam S1021 pump system equipped
with UV variable wavelength Knauer and GE radiodetectors.
were found to be À7.763, À6.874, and À5.946 for MefWAY,
MPPF, and WAY-100635, respectively. These data directly corre-
late the high binding potential of BMPPSiF with the bivalent
binding mode obtained in the biological studies.
1H NMR, 13C NMR, 19F NMR, and 29Si NMR spectra were recorded on
Bruker DPX-200 FT (1H: 200 MHz, 13C: 50.2 MHz), Bruker AVANCE-
300 FT (1H: 300 MHz, 13C: 75.5 MHz), and Bruker DPX-400 FT (1H:
400 MHz, 13C: 100.2 MHz) spectrometers by using the indicated in-
ternal reference. The chemical shifts (d) and coupling constants (J)
are expressed in ppm and Hz, respectively.
Conclusions
In conclusion, the purpose of this study was to develop a valua-
ble prosthetic group that can be used in direct and indirect
18F-labeling methods, which opens the route to produce
a wide array of dimeric-specific PET tracers. Although dimeric,
trimeric, and multimeric radioligands have been labeled with
[18F] atoms, prosthetic groups to prepare dimeric ligands have
not been offered to date, and there are several methodical and
salient advantages in our approach. Our method with SiX–di-
propargyl glycerol (X: H, F, or 18F) provides technical advantag-
es for the preparation of dimeric ligands by prelabeling or
post-labeling. Thus, a bivalent homodimeric SiFA-derivatized
radioligand, [18F]BMPPSiF, with enhanced affinity was devel-
oped as a model to show its application as a neuroimaging
PET tracer with preferential uptake in 5-HT1A receptor rich re-
gions. Homology modeling and molecular docking studies of
BMPPSiF and known antagonists (WAY-100635, MPPF, and
MefWAY) with dimeric and multimeric 5-HT1A receptor models
were demonstrated, with the highest G score obtained for the
newly synthesized BMPPSiF compared with those of the other
antagonists.
Mass spectra were recorded on a Nermag R10-10C apparatus.
High-resolution mass spectra were performed by the CESAMO (Tal-
ence, France) and were recorded on a Q-TOF tadem mass spec-
trometer (API Q-STAR Pulsar i, Applied Biosystems). Positive-ion-
mode ESI mass spectrometry was used for the analysis. MALDI-MS
data were collected with a CESAMO (Talence, France) device on
a Voyager mass spectrometer (Applied Biosystems). The instrument
was equipped with a pulsed N2 laser (337 nm) and a time-delayed
extracted ion source. Melting points were determined by using
a Bꢁchi–Totolli apparatus and a Stuart Scientific apparatus (SMP3)
and are uncorrected.
In vivo imaging was performed by using micro-PET imaging as
noninvasive technology. The acquisitions were performed by using
a GE FLEX Triumph micro-PET/SPECT/CT scanner. This state-of-the-
art scanner consisted of a micro-PET module (LabPET4) with 2ꢂ2ꢂ
10 mm3 LYSO/LGSO scintillators in an 8-pixel, quad-APD detector
module arrangement. The micro-CT part consisted of a high-resolu-
tion micro-CT tube with a focal spot size that was switchable be-
tween 10 or 50 mm.
Experimental Section
Syntheses
General
4-(Di-tert-butylsilanyl)phenol (1): An anhydrous THF (10 mL) solu-
tion of 4-bromophenol (3.46 g, 20 mmol, 1 equiv) was added to
a stirred suspension of sodium hydride (60% dispersion in mineral
oil; 880 mg, 22 mmol, 1.1 equiv) in anhydrous THF (90 mL). The
mixture was heated at 558C for 3 h and then cooled to À658C. A
1.6m solution of tert-butyllithium (26.3 mL, 42 mmol, 2.1 equiv)
was added to the reaction mixture. After 30 min, di-tert-butylchlor-
osilane (4.25 mL, 21 mmol, 1.05 equiv) was added, and the reaction
was warmed to room temperature and stirred overnight. The re-
sulting mixture was quenched with saturated NH4Cl (100 mL) and
extracted with Et2O (3ꢂ150 mL). The resulting organic layer was
successively washed with water (100 mL) and brine (100 mL), dried
over Na2SO4, and filtered, then the solvent was removed in vacuo.
The crude product was purified by column chromatography (silica
gel, petroleum ether/ethyl acetate 95:5 to 9:1) to afford the de-
sired compound as a colorless solid (3.7 g, 78%); mp: 85–878C;
1H NMR (300 MHz, CDCl3): d=7.45 (d, 2H, J=7.5 Hz), 6.82 (d, 2H,
J=7.5 Hz), 4.91 (s, 1H), 3.83 (s, 1H), 1.03 ppm (s, 18H); 13C NMR
(75 MHz, CDCl3): d=156.4, 137.5, 126.7, 114.9, 29.1, 19.2 ppm;
29Si NMR (60 MHz, CDCl3): d=13.17 ppm; HRMS (ESI): m/z calcd for
C14H24OSiNa: 259.1488 [M+Na]+; found: 259.1485.
All water-sensitive reactions were carried out under a nitrogen at-
mosphere with dry solvents under anhydrous conditions. Yields
refer to chromatographically and spectroscopically (1H NMR) homo-
geneous materials. Commercial reagents were used without purifi-
cation, unless otherwise stated. Macherey–Nagel silica gel 60M
(230–400 mesh ASTM) was used for flash chromatography. The
boiling range for petroleum ether is 40–608C. Et3N was distilled
over CaH2. THF and Et2O were distilled from sodium and benzo-
phenone. Ethanol and methanol were dried over magnesium turn-
ings activated by iodine. In some cases, THF, Et2O, CH2Cl2, and
MeOH, were dried with a MB SPS-800 purification system. Micro-
wave-assisted reactions were carried out on a Biotage initiator.
Spiperone, serotonin, and ketanserin (+)-tartrate salt were pro-
cured from Sigma–Aldrich; paroxetine hydrochloride was obtained
from Fluka. All solvents used were of analytical grade. TLC was run
on silica gel coated aluminum sheets (silica gel 60 F254; Merck,
Germany) and visualized in UV light (254 nm). Radiocomplexation
and radiochemical purity were checked by radio-TLC.
High-performance liquid chromatography (HPLC) analyses were
performed by using a Phenomenex Luna C18 column (250ꢂ
4.6 mm, 5 mm, 1 mLminÀ1) under the indicated conditions. Analyti-
cal HPLC systems used were either a JASCO system with Chrom-
NAV software, a PU-2089 Plus quaternary gradient pump, and
a MD-2010 Plus photodiode array detector or a Dionex P680
system with Chromeleon software and a Dionex UVD170U system,
equipped with UV multiwavelength and Raytest Gabi Star detec-
tors.
Di-tert-butyl(4-((2,2-dimethyl-1,3-dioxolan-4-yl)methoxy)phenyl)-
silane (2): DEAD (6.97 g, 6.3 mL, 40 mmol, 2.5 equiv) was slowly
added at room temperature to a stirred solution of silane (3.78 g,
16 mmol, 1 equiv), solketal (4.23 g, 3.99 mL, 32 mmol, 2 equiv), and
PPh3 (10.5 g, 40 mmol, 2.5 equiv) in anhydrous THF (100 mL). The
mixture was then heated to reflux overnight. After one night, sol-
ketal (2.11 g, 2.0 mL, 16 mmol), PPh3 (4.2 g, 16 mmol), and DEAD
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 345

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
(2.79 g, 2.5 mL, 16 mmol) were added successively at room temper-
ature, and the reaction mixture was then heated at reflux for an-
other 24 h. The resulting mixture was evaporated in vacuo, and
pentane (200 mL) was added to the slurry to precipitate Ph3PO.
Ph3PO was then removed by filtration, and the resulting mixture
was concentrated in vacuo. The crude product was purified by
column chromatography (silica gel, petroleum ether/ethyl acetate
95:5) to afford the desired compound as a colorless solid (3.55 g,
152.4, 141.3, 123.1, 121.1, 118.3, 111.3, 59.4, 57.9, 55.5, 53.2,
50.9 ppm; HRMS (ESI): m/z calcd for C13H21N2O2: 237.1597 [M+H]+;
found: 237.1595.
1-(2-Azidoethyl)-4-(2-methoxyphenyl)piperazine
(2’):
DBU
(457 mg, 448 mL, 3 mmol, 2 equiv) and DPPA (825 mg, 647 mL,
3 mmol, 2 equiv) were added to a solution of 1-(2-hydroxyethyl)-4-
(2-methoxyphenyl)piperazine (1’; 354 mg, 1.5 mmol, 1 equiv) in an-
hydrous DMF (3.75 mL) under argon. The reaction mixture was
stirred at 658C overnight. After that time, the reaction mixture was
cooled to room temperature, and water (10 mL) was added. The
mixture was then extracted with Et2O (2ꢂ10 mL). The resulting or-
ganic layer was successively washed with water (10 mL) and brine
(10 mL), dried over Na2SO4, and filtered, then the solvent was re-
moved in vacuo. The crude product was purified by column chro-
matography (silica gel, petroleum ether/ethyl acetate 1:1) to afford
the desired compound as a yellow oil (360 mg, 91%). 1H NMR
(300 MHz, CDCl3): d=6.82–7.05 (m, 4H), 3.86 (s, 3H), 3.40 (t, 2H,
J=6.1 Hz), 3.11 (m, 4H), 2.60–2.75 ppm (m, 6H); 13C NMR (75 MHz,
CDCl3): d=152.4, 141.3, 123.1, 121.1, 118.4, 111.3, 57.3, 55.5, 53.5,
50.7, 48.4 ppm; HRMS (ESI): m/z calcd for C13H20N5O: 262.1662 [M+
H]+; found: 262.1670.
1
63%); mp: 55–588C; H NMR (300 MHz, CDCl3): d=7.49 (d, 2H, J=
8.5 Hz), 6.90 (d, 2H, J=8.5 Hz), 4.49 (m, 1H), 4.18 (m, 1H), 4.07 (dd,
1H, J=9.4, 5.4 Hz), 3.93 (m, 2H), 3.83 (s, 1H), 1.47 (s, 3H), 1.41 (s,
3H), 1.03 ppm (s, 18H); 13C NMR (75 MHz, CDCl3): d=159.4, 137.3,
127.0, 113.9, 109.9, 74.1, 68.5, 67.1, 29.1, 26.9, 25.5, 19.2 ppm;
29Si NMR (60 MHz, CDCl3): d=13.08 ppm; HRMS (ESI): m/z calcd for
C20H34O3SiAg: 457.1322 [M+Ag]+; found: 457.1318.
3-(4-(Di-tert-butylsilyl)phenoxy)propane-1,2-diol (3): The protect-
ed compound (2.7 g, 7.7 mmol, 1 equiv) was dissolved in MeOH
(27 mL). Dowex 50WX8-200 resin (2.7 g) was then added, and the
reaction mixture was stirred at room temperature for 48 h. The
resin was removed by filtration, and the MeOH was removed in
vacuo. The crude product was purified by column chromatography
(silica gel, petroleum ether/ethyl acetate 2:8) to afford the desired
1
4,4’-(2,2’-(4,4’-(3-(4-(di-tert-butylsilyl)phenoxy)propane-1,2-
diyl)bis(oxy)bis(methylene)bis(1H-1,2,3-triazole-4,1-diyl))bis-
(ethane-2,1-diyl))bis(1-(2-methoxyphenyl)piperazine) (5): In a mi-
crowave vial, (4-(2,3-bis(prop-2-ynyloxy)propoxy)phenyl)di-tert-bu-
tylsilane (4; 97 mg, 0.25 mmol, 1.0 equiv), 1-(2-azidoethyl)-4-(2-me-
thoxyphenyl)piperazine (2’; 131 mg, 0.5 mmol, 2 equiv), sodium as-
corbate (59 mg, 0.3 mmol, 1.2 equiv), and copper sulfate (16 mg,
0.1 mmol, 0.4 equiv) were suspended in a mixture of tert-butanol/
water (3:1, 1.2 mL). The reaction mixture was heated at 1008C
under microwaves (50 W) for 15 min. After cooling of the reaction,
water (5 mL) was added, and the solution was extracted with
AcOEt (2ꢂ5 mL). The organic layers were then dried over Na2SO4.
After filtration, the solvent was removed in vacuo, and the crude
product was purified by column chromatography (silica gel, AcOEt
to AcOEt/MeOH 95:5) to afford the desired compound as a white
foam (160 mg, 70%). Analytical HPLC method: Phenomenex Luna
C18 column (250ꢂ4.6 mm, 5 mm), linear gradient 90:10 A (0.1%
TFA in water)/B (MeCN) to 10:90 A/B over 30 min at a flow rate of
1 mLminÀ1, Rt =22.3 min; 1H NMR (400 MHz, CDCl3): d=7.77 (s,
1H), 7.73 (s, 1H), 7.45 (d, 2H, J=8.0 Hz), 6.81–7.06 (m, 10H), 4.87
(s, 2H), 4.70 (s, 2H), 4.47 (t, 4H, J=6.4 Hz), 4.00–4.13 (m, 3H), 3.86
(s, 6H), 3.81 (s, 1H), 3.71–3.80 (m, 2H), 3.01–3.15 (m, 8H), 2.88 (t,
4H, J=6.4 Hz), 2.63–2.75 (m, 8H), 1.02 ppm (s, 18H); 13C NMR
(100 MHz, CDCl3): d=159.5, 152.4, 145.3, 144.9, 141.2, 137.3, 126.8,
123.6, 123.4, 123.2, 121.1, 118.4, 113.9, 111.3, 76.7, 70.1, 67.4, 65.1,
64.2, 57.7, 55.5, 53.5, 50.7, 47.7, 29.1, 19.2 ppm;29Si NMR (80 MHz,
CDCl3): d=12.50 ppm; HRMS (ESI): m/z calcd for C49H73N10O5Si:
909.5529 [M+H]+; found: 909.5534.
compound as a colorless solid (2.3 g, 96%); mp: 77–798C; H NMR
(300 MHz, CDCl3): d=7.49 (d, 2H, J=8.6 Hz), 6.90 (d, 2H, J=
8.6 Hz), 4.02–4.14 (m, 3H), 3.70–3.90 (m, 3H), 1.03 ppm (s, 18H);
13C NMR (75 MHz, CDCl3): d=159.3, 137.3, 127.2, 113.9, 70.7, 68.8,
63.9, 29.0, 19.1 ppm; 29Si NMR (60 MHz, CDCl3): d=13.07 ppm;
HRMS (ESI): m/z calcd for C17H30O3SiAg: 417.1013 [M+Ag]+; found:
417.1009.
(4-(2,3-Bis(prop-2-ynyloxy)propoxy)phenyl)di-tert-butylsilane (4):
NaH (200 mg, 8.31 mmol, 3 equiv) was suspended in anhydrous
THF (28 mL). The diol (860 mg, 2.77 mmol, 1 equiv) was slowly
added to the reaction mixture at 08C, and the mixture was re-
fluxed for 4 h. The reaction mixture was cooled to 08C, and prop-
argyl bromide (80% in toluene; 988 mg, 895 mL, 8.31 mmol,
3 equiv) was added. The reaction mixture was then held at reflux
overnight. The resulting mixture was cooled to room temperature
and quenched with water (20 mL). Et2O (50 mL) was added, and
the resulting organic layer was successively washed with water
(20 mL) and brine (50 mL), dried over Na2SO4, and filtered. The sol-
vent was removed in vacuo. The crude product was purified by
column chromatography (silica gel, petroleum ether/ethyl acetate
95:5) to afford the desired compound as a yellow oil (870 mg,
1
81%). H NMR (300 MHz, CDCl3): d=7.48 (d, 2H, J=7.5 Hz), 6.90 (d,
2H, J=7.5 Hz), 4.39 (d, 2H, J=2.4 Hz), 4.21 (d, 2H, J=2.4 Hz),
4.07–4.15 (m, 3H), 3.72–3.85 (m, 3H), 2.44 (2t, 6H, J=2.4 Hz),
1.03 ppm (s, 18H); 13C NMR (75 MHz, CDCl3): d=159.4, 137.2, 126.8,
114.0, 80.0, 79.5, 75.9, 74.9, 74.7, 69.6, 67.4, 58.8, 58.0, 29.1,
19.2 ppm; 29Si NMR (60 MHz, CDCl3): d=12.52 ppm; HRMS (ESI): m/
z calcd for C23H34O3SiNa: 409.2169 [M+Na]+; found: 409.2172.
Synthesis of 4,4’-(2,2’-(4,4’-(3-(4-(di-tert-butylfluorosilyl)phen-
oxy)propane-1,2-diyl)bis(oxy)bis(methylene)bis(1H-1,2,3-triazole-
4,1-diyl))bis(ethane-2,1-diyl))bis(1-(2-methoxyphenyl)piperazine)
(6): Potassium fluoride (7 mg, 0.117 mmol, 2 equiv), Kryptofix[2.2.2]
(44 mg, 0.117 mmol, 2 equiv), and acetic acid (11 mg, 11 mL,
0.117 mmol, 2 equiv) were added to a solution of silane 5 (53 mg,
0.058 mmol, 1 equiv) in anhydrous DMSO (2.5 mL). The reaction
mixture was heated at 808C overnight. The reaction mixture was
cooled to room temperature and diluted with water (5 mL) and
Et2O (5 mL). The organic layer was washed with brine (5 mL) and
dried over MgSO4, then the solvent was removed in vacuo to
afford the desired compound (49 mg, 91%), which was then used
as an HPLC reference and also for isotope exchange 18F labeling.
1-(2-Hydroxyethyl)-4-(2-methoxyphenyl)piperazine (1’): 2-Bro-
moethanol (1.15 g, 650 mL, 9.18 mmol, 1.05 equiv) was added to
a
solution of 1-(2-methoxyphenyl)piperazine hydrochloride in
Scheme 2 (2 g, 8.74 mmol, 1 equiv) and anhydrous potassium car-
bonate (3 g, 21.85 mmol, 2.5 equiv) in anhydrous acetonitrile
(20 mL). The reaction mixture was heated at reflux for 48 h. After
that time, the mixture was filtered and concentrated. The resulting
slurry was then purified by column chromatography (silica gel,
CHCl3/MeOH 9:1) to afford the desired compound as a colorless
1
solid (2 g, 97%); mp: 70–728C; H NMR (300 MHz, CDCl3): d=6.82–
7.05 (m, 4H), 3.86 (s, 3H), 3.65 (t, 2H, J=5.4 Hz), 3.10 (m, 4H), 2.72
(m, 4H), 2.62 ppm (t, 2H, J=5.4 Hz); 13C NMR (75 MHz, CDCl3): d=
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 346

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Analytical HPLC: method A: Phenomenex Luna C18 column (250ꢂ
4.6 mm, 5 mm), linear gradient 90:10 A (0.1% TFA in water)/B
(MeCN) to 10:90 A/B over 30 min at a flow rate of 1 mLminÀ1, Rt =
21.5 min; method B: Phenomenex Luna C18 column (250ꢂ4.6 mm,
5 mm), isocratic gradient 53:47 A (0.1% TFA in water)/B (MeCN) at
pure [18F]BMPPSiF to the initial radioactivity of [18F]-fluoride (at the
end of bombardment).
Measurement in hippocampal primary cultures
1
a flow rate of 1 mLminÀ1, Rt =17.9 min; H NMR (400 MHz, CDCl3):
Primary cultured cortical neuronal cells were obtained from the
hippocampus of fetal Sprague–Dawley rats at 17–20 days of gesta-
tion, according to a previously described procedure. In brief, single
cells dissociated from the hippocampus of fetal rats were sited on
the poly-l-lysine coated culture slide at 3ꢂ105 cells per well. The
culture was incubated in Dulbecco’s modified Eagle’s medium (F-
12) supplemented with 10% heat-inactivated fetal bovine serum
for 1–7 days after plating or with 10% heat-inactivated horse
serum for 8–15 days after plating, in each case together with
2.5 mm glutamine, 17.5 mm glucose, and 14.3 mm NaHCO3. The
cultures were subsequently maintained at 378C in a humidified
5% CO2 atmosphere. Three days subsequent to plating, the non-
neuronal cells were removed by addition of cytosine arabinoside
at 10 mm concentration. Only mature cultures surviving 15–18 days
in vitro were used for the experiments. In the cell culture, the ratio
of nerve to glial was found to be approximately 80:20. After
18 days of culture in vitro, successfully cultured neurons were se-
lected for binding studies, which were carried out. The amount of
radioactivity (counts per minute) in cell lysates was determined by
gamma scintillation counting. The uptake was then calculated and
d=7.77 (s, 1H), 7.74 (s, 1H), 7.49 (d, 2H, J=8.4 Hz), 6.83–7.05 (m,
10H), 4.87 (s, 2H), 4.69 (s, 2H), 4.47 (t, 4H, J=6.4 Hz), 4.01–4.15 (m,
3H), 3.85 (s, 6H), 3.70–3.82 (m, 2H), 2.99–3.16 (m, 8H), 2.88 (t, 4H,
J=6.4 Hz), 2.63–2.78 (m, 8H), 1.03 ppm (s, 18H); 13C NMR
(100 MHz, CDCl3): d=159.9, 152.4, 145.2, 144.9, 141.2, 135.6, 134.9
(d, 2JCÀF =14.1 Hz), 123.6, 123.4, 123.2, 121.1, 118.3, 114.1, 111.3,
76.7, 70.0, 67.5, 65.1, 64.2, 57.7, 55.5, 53.5, 50.7, 47.7, 27.5,
20.4 ppm (d, 2JCÀF =12.1 Hz); 19F NMR (377 MHz, CDCl3): d=
1
À189.6 ppm; 29Si NMR (80 MHz, CDCl3): d=14.37 ppm (d, JSiÀF
=
296.5 Hz); HRMS (ESI): m/z calcd for C49H72FN10O5Si: 927.5434 [M+
H]+; found: 927.5457.
Cytotoxicity assay
For the cell viability assay, an optimized number of cells (6000 cells
per well in 96-well plates) were seeded into the wells of a flat-bot-
tomed 96-well microtiter plate in triplicate 12 h before the addition
of serial dilutions of 5 or 6 in the range of 0.1 nm–10 mm. The
plates were developed by treatment with XTT dye (Sigma) 2 h and
4 h after the addition of 5 or 6.
expressed as pmol(mg protein)À1 minÀ1
.
Competitive binding with unlabeled serotonin was assessed, and
the dissociation constant was calculated. Final concentrations in
the well were 0.01–10000 nm. Hippocampal cultures were washed
with Hanks’ balanced salt solution (HBSS) and were then incubated
for 20 min in HBSS at 378C prior to the experiment. Binding experi-
ments was conducted at 378C. Cells were incubated for 30 min
with increasing concentrations (0.001 nm–1 mm) of [18F]BMPPSiF in
the absence and presence of a 100-fold excess of unlabeled seroto-
nin to estimate the total binding and nonspecific binding, respec-
tively. Specific binding was obtained by subtracting the nonspecific
binding from the total binding. At the end of each experiment, the
cells were washed with cold phosphate-buffered saline and 0.9%
saline four times. The cell-associated radioactivity was determined
by scintillation counting calibrated for 511 KeV.
18F Radiolabeling
All fluorination experiments were run on a commercial TRACERLab
FxFÀN (Ge Healthcare) instrument. No-carrier-added [18F]-fluoride
was produced through the 18O(p,n)18F nuclear reaction by irradia-
tion of enriched [18O]-H2O. [18F]-Fluoride production: bombardment
80–100 min; 40 mA; target volume 1.8–2.5 mL; 100–130 GBq are
obtained. [18F]-Fluoride was trapped on an anion-exchange resin
cartridge (Sep-Pak QMA light, Waters). The cartridge was eluted
with a solution of Kryptofix[2.2.2] (22 mg) and potassium carbonate
(7 mg) in H2O (0.3 mL) and MeCN (0.3 mL). The solvents were re-
moved by heating at 1008C for 8 min under a gentle stream of ni-
trogen. During this time, MeCN (4ꢂ0.25 mL) was added and
evaporated to give the dry K[18F]/Kryptofix[2.2.2] complex. The
peak of the 18F-labeled products was confirmed by comparison
with the HPLC retention times of the nonradioactive reference
molecule and the precursor, under both isocratic and linear gradi-
ent conditions.
5-HT1A, 5-HT2A, D2, and 5-HTT receptor binding assays of
[18F]BMPPSiF
All animal experiments were conducted in accordance with the
guidelines of the INMAS animal ethics committee (CPCSEA Regn
No.8/GO/a/99). Different parts of rat brains were homogenized and
prepared as described elsewhere for the determination of bioactiv-
ity.
[18F]BMPPSiF method: A solution of 5 or 6 (6 mg) and glacial
acetic acid (3 mL) in anhydrous DMSO (300 mL) was added to the
dry K[18F]/Kryptofix[2.2.2] complex. After being heated at 808C for
15 min, the reaction mixture was diluted with MeCN (2.25 mL) and
TFA (0.1% in water, 2.75 mL). This solution was injected onto a sem-
ipreparative HPLC Phenomenex Luna C18 column (250ꢂ10 mm,
5 mm) and eluted with MeCN/0.1% TFA in water (45:55) at a flow
rate of 2.5 mLminÀ1. The Rt value was 13 min. The isolated fluori-
nated compound was diluted into water (40 mL) and collected
onto a C18 Sep-Pak cartridge. The cartridge was washed with water
(10 mL) and eluted with ethanol (2 mL). The compound was then
formulated in an aqueous physiological solution (0.9% NaCl solu-
tion), and the purity was checked by analytical HPLC on a Phenom-
enex Luna C18 column (250ꢂ4.6 mm, 5 mm) with elution with
MeCN/0.1% TFA in water (47:53) at a flow rate of 1 mLminÀ1. The
Rt value was 6.8 min. The radiochemical yield (30%, not decay cor-
rected, n=4) was calculated as the ratio of the radioactivity of
5-HT1A: The hippocampus of the rat brain was homogenized in
10 volumes of ice-cold buffer (50 mm tris(hydroxymethyl)aminome-
thane–HCl (Tris-HCl) pH 7.6) by using an Ultra Turrax T10 apparatus
(IKA). The homogenate was centrifuged at 20000 g for 10 min. The
resulting pellet was resuspended with the Ultra Turrax apparatus
and centrifuged again at 20000 g for 10 min. The same procedure
was repeated again. The pellet was then resuspended in 10 vol-
umes of buffer and stored at À808C until used in binding studies.
The binding assay was carried out in a final volume of 2.5 mL Tris-
HCl buffer (50 mm, pH 7.4, 0.1% ascorbic acid, 2 mm CaCl2) con-
taining membrane homogenate (~0.5 mgmLÀ1 protein) and vari-
ous concentrations of [18F]BMPPSiF. Nonspecific binding was de-
fined as the amount of [18F]BMPPSiF bound in the presence of
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 347

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
10 mm serotonin. The binding assay was performed in triplicate at
208C for 120 min. The incubation was terminated by rapid filtration
through GF/B glass fiber filters. The filters were rapidly washed
with 4 mL portions of ice-cold buffer, transferred into 4 mL of scin-
tillation cocktail, and analyzed for radioactivity by scintillation
counting.
Biodistribution
In the biodistribution study, an intravenous injection of
[18F]BMPPSiF conjugate in volume of 100 mL (5 nmolkgÀ1
,
a
7.2 MBq activity) was performed through the tail vein of each
mouse (n=5 for each group). At 10 min, 30 min, 1 h, 2 h, and 4 h
post-injection, mice were dissected to take out different tissues,
which were weighed and counted in a multichannel analyzer cali-
brated for 511 KeV energy. Uptake of the radiotracer in each tissue
was calculated and expressed as a percentage injected dose per
gram of the tissue (% IDgÀ1). For blocking studies, rats received
a tail-vein injection of a solution (100 mL) of 8-hydroxy-N,N-diprop-
yl-2-aminotetralin (1 mgkgÀ1), a full agonist of the 5-HT1A receptor,
5 min before administration of the radiotracer and were then dis-
sected. The brain was removed, chilled, and dissected at 10, 30,
and 60 min. Samples from different brain regions (cortex, hippo-
campus, cerebellum, etc.) were collected, weighed, and counted.
Regional biodistribution was performed by dissecting the different
parts of the rat brain.
5-HT2A: The cortex of the rat brain was homogenized and prepared
analogously to the procedure described above and stored at
À808C until used in binding studies. The assay was performed in
a 5.0 mL volume of Tris-HCl buffer (pH 7.6) containing a 100-fold
excess of ketanserin tartarate.
D2: The striatum of the rat brain was homogenized and prepared
analogously to the procedure described above and stored at
À808C until used in binding studies. The final 5.0 mL volume of
the binding assay contained Tris-HCl buffer (pH 7.4) with 10 mm
MgCl2, 1 mm ethylenediaminetetraacetate (EDTA), membrane ho-
mogenate (~0.2 mgmLÀ1 protein), and various concentrations of
[18F]BMPPSiF, which were dissolved and diluted with buffer as de-
scribed above. Nonspecific binding was determined with a 100-
fold excess of spiperone.
5-HT transporter: The caudate nucleus of the rat brain was homo-
genized and prepared analogously to the procedure described
above and stored at À808C until used in binding studies. The final
5.0 mL volume of the binding assay contained Tris-HCl buffer
(pH 7.4) with 120 mm NaCl, 5 mm KCl, membrane homogenate (~
1.2 mgmLÀ1 protein), and various concentrations of [18F]BMPPSiF in
the presence and absence of a 100-fold excess of proxetine. Filtra-
tion and counting of the samples were the same as described
above.
Data analysis
The competition curve of the receptor binding experiments was
analyzed by nonlinear regression by using the algorithms in Graph-
Pad PRISM 5.0 (San Diego, CA). The competitive curve was fitted to
a sigmoid curve, in which the Kd value and the Hill coefficient were
free parameters. Biodistribution data are represented as the mean
Æstandard deviation.
Computational studies: All of the computational studies were per-
formed with the molecular modeling software from Schrçdin-
ger LLC, New York, NY, 2013: Maestro 9.5, Prime Version 2.3, Lig-
prep Version 2.9, and Glide Version 6.
PET and CT acquisition
Animals were injected with 37 MBq of [18F]BMPPSiF immediately
prior to their micro-PET scan on the camera bed at the start of a dy-
namic acquisition. Frames of 1ꢂ1 and 5ꢂ5 min were accordingly
sequentially recorded for 60 min. For anatomical localization,
a micro-CT scan was sequentially acquired by using 512 projections
over 3608 at 70 kVp/180 mA and two magnifications with a spot
size of 50 mm. The resulting PET was reconstructed by using 50
iterations of the Maximum Likelihood Expectation Maximization al-
gorithm. CT reconstruction was done by using Triumph. All images
were reconstructed and analyzed with VIVID (Amira, San Diego,
USA). Time–activity curves and ROI analyses was done by using
a comprehensive set of models including compartment, noncom-
partment, and reference models with General Kinetic Modeling
Tool (PKIN), Version 3.205 (PMOD Technologies Ltd., Sumatrastrasse,
Zꢁrich).
Homology modeling: The human 5-HT1A sequence was retrieved
from the NCBI database (accession code: NP_000515.2). This se-
quence was used for the modeling of 5-HT1A by using the Prime
homology modeling tool of Schrçdinger. By using Prime, three ho-
mology models were prepared, monomeric, dimeric, and multimer-
ic models, based on the crystal structure of the b2-adrenergic re-
ceptor as a template in the absence of cholesterol.
The preliminary monomer model of 5-HT1A was built by using the
crystal structure of the b2-adrenergic receptor (PDB code: 2RH1),
with a resolution of 2.4 ꢃ.[31] Sequence alignment was performed
by using Clustal W under the GPCR-specific mode. Secondary struc-
ture prediction, supported with PSIPRED, was performed, and this
is best known for optimal predictions. A 2RH1-aligned 5-HT1A
model was obtained by building backbone atom coordinates for
aligned regions and side chains of conserved residues, optimiza-
tion of side chains, minimization of nontemplate residues, building
of insertions, and closing of deletions in the alignment. Finally, the
structure was refined, and protein preparation was performed by
using the Protein Preparation Wizard of Schrçdinger.
Preparation of animal models with neurological deficits
For 5-HT/D2 depletion rat models, p-chloroamphetamine (PCA;
8 mgkgÀ1, intraperitoneal), a serotonin neurotoxin, is a useful tool
for chemical lesioning of brain serotonin neurons. The serotonin-
depletion SD rat model was created by injection of a PCA saline
solution (10 mgkgÀ1) into the peritoneal cavity; this was a dose
sufficient to produce long-term depletion of brain serotonin. After
21 days, 5-HT1A receptor mRNA expression was increased by 80%
in the cortex and decreased by 50% in the hippocampus, as previ-
ously reported.[30] Thus, the experiment was conducted 21 days
after PCA administration.
The 5-HT1A receptor tends to form dimeric and multimeric models,
as validated by experimental approaches.[32] Hence, we have vali-
dated our results by performing docking studies on dimeric and
multimeric models. For the dimeric models, the transmembrane re-
gions TM4 and TM5 of the monomer 5-HT1A were superimposed
on rhodopsin as a template (PDB code: 1N3M). Only TM helices
were superimposed, that is, chains A and C of 1N3M. The final di-
meric structure was optimized and refined by removing steric
clashes. Similarly, for the multimeric model, the monomer model of
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 348

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
K. Jurkschat, Angew. Chem. 2006, 118, 6193–6197; Angew. Chem. Int. Ed.
2006, 45, 6047–6050.
[11] C. Wꢄngler, A. Kostikov, J. Zhu, J. Chin, B. Wꢄngler, R. Shirrmacher, Appl.
Sci. 2012, 2, 277–302.
[12] W. J. McBride, C. A. D’Souza, R. M. Sharkey, H. Karacay, E. A. Rossi, C. H.
Chang, D. M. Goldenberg, Bioconjugate Chem. 2010, 21, 1331–1340.
[13] W. J. McBride, R. M. Sharkey, D. M. Goldenberg, EJNMMI Res. 2013, 3, 36.
[14] C. Wꢄngler, S. Niedermoser, J. Chin, K. Orchowski, E. Schirrmacher, K. Ju-
rkschat, L. Iovkova-Berends, A. P. Kostikov, R. Schirrmacher, B. Wꢄngler,
Nat. Protoc. 2012, 7, 1946–1955.
[15] A. P. Kostikov, J. Chin, K. Orchowski, E. Schirrmacher, S. Niedermoser, K.
Jurkschat, L. Iovkova-Berends, C. Wꢄngler, B. Wꢄngler, R. Schirrmacher,
Bioconjugate Chem. 2012, 23, 106–114.
the 5-HT1A receptor was aligned with the template 1N3M, and
a multimeric 5-HT1A model with six chains was prepared.
Binding site prediction: For all three models, grid calculations
were performed for the protein active site as per the residues that
are known to be important for ligand binding from residue muta-
genesis studies in the case of the monomer 5-HT1A. Hence, a grid
was developed of the important antagonist binding site residues.
For the dimeric and multimeric models, the binding pockets were
identified by using the Sitemap tool of Schrçdinger, which scored
the five most prominent binding sites and arranged them accord-
ing to the scores. All sites were used for grid generation for the di-
meric and multimeric models, and ligands were docked at all five
sites.
[16] B. Wꢄngler, A. P. Kostikov, S. Niedermoser, J. Chin, K. Orchowski, E.
Schirrmacher, L. Iovkova-Berends, K. Jurkschat, C. Wꢄngler, R. Schirr-
macher, Nat. Protoc. 2012, 7, 1964–1969.
Molecular docking: Glide, Version 6, was used for docking analysis,
which predicts the binding mode of the ligand with high accuracy.
The docking study was initiated by bringing specified prepared
proteins and ligand molecules together. Glide docking in the XP
mode was performed for the monomer binding site and all five
sites for both the dimeric and multimeric models. This process
evaluates the energy interactions of the ligand with the protein in
terms of a G score value, which is used for predicting binding affin-
ity.
[17] J. Schulz, D. Vimont, T. Bordenave, D. James, J.-M. Escudier, M. Allard, M.
Szlosek-Pinaud, E. Fouquet, Chem. Eur. J. 2011, 17, 3096–3100.
[18] E. Amigues, J. Schulz, M. Szlosek-Pinaud, P. Fernandez, S. Silvente-
Poirot, S. Brillouet, F. Courbon, E. Fouquet, ChemPlusChem 2012, 77,
345–349.
[19] M. Bai, Cell. Signalling 2004, 16, 175–186.
[20] S. Butini, G. Campiani, S. Franceschini, F. Trotta, V. Kumar, E. Guarino, G.
Borrelli, I. Fiorini, E. Novellino, C. Fattorusso, M. Persico, N. Orteca, K.
Sandager-Nielsen, T. A. Jacobsen, K. Madsen, J. Scheel-Kruger, S.
Gemma, J. Med. Chem. 2010, 53, 4803–4807.
[21] J. Sandell, C. Halldin, V. Pike, Y. Chou, K. Varnas, H. Hall, S. Marchais, B.
Nowicki, H. Wikstrom, C. Swahn, L. Farde, Nucl. Med. Biol. 2001, 28,
177–185.
Acknowledgements
[22] M. Karramkam, F. Hinnen, M. Berrehouma, C. Hlavacek, F. Vaufrey, C.
Halldin, J. A. McCarron, V. W. Pike, F. Dolle, Bioorg. Med. Chem. 2003, 11,
2769–2782.
[23] J. A. McCarron, V. W. Pike, C. Halldin, J. Sandell, J. Sovago, B. Gulyas, Z.
Cselenyi, H. V. Wikstrom, S. Marchais-Oberwinkler, B. Nowicki, F. Dolle, L.
Farde, Mol. Imaging Biol. 2004, 6, 17–26.
The work was supported by the Defence Research and Develop-
ment Organization, Ministry of Defence (India), under R&D proj-
ect INM-311(3.1).
[24] N. Saigal, R. Pichika, B. Easwaramoorthy, D. Collins, B. T. Christian, B. Shi,
T. K. Narayanan, S. G. Potkin, J. Mukherjee, J. Nucl. Med. 2006, 47, 1697–
1706.
Keywords: click chemistry
radiolabeling · silicon
· fluorides · ligand design ·
[25] T. Bordenave, P. P. Hazari, D. James, A. K. Mishra, M. Szlosek-Pinaud, E.
Fouquet, Eur. J. Org. Chem. 2013, 1214–1217.
[1] H. S. Gill, J. Marik, Nat. Protoc. 2011, 6, 1718–1725.
[26] S. M. Sternson, J. C. Wong, C. M. Grozinger, S. L. Schreiber, Org Lett.
2001, 3, 4239–4242.
[27] D. James, J.-M. Escudier, E. Amigues, J. Schulz, C. Vitry, T. Bordenave, M.
Szlosek-Pinaud, E. Fouquet, Tetrahedron Lett. 2010, 51, 1230–1232.
[28] N. Singh, P. P. Hazari, S. Prakash, K. Chuttani, H. Khurana, H. Chandra,
A. K. Mishra, Med. Chem. Commun. 2012, 3, 814–823.
[29] J. Kꢁhhorn, H. Hꢁbner, P. Gmeiner, J. Med. Chem. 2011, 54, 4896–4903.
[30] A. Garcꢅa-Osta, D. Frechilla, J. Del Rꢅo, J. Neurochem. 2000, 74, 1790–
1797.
[31] V. Cherezov, D. M. Rosenbaum, M. A. Hanson, S. G. F. Rasmussen, F. S.
Thian, T. S. Kobilka, H.-J. Choi, P. Kuhn, W. I. Weis, B. K. Kobilka, R. C. Ste-
vens, Science 2007, 318, 1258–1265.
[2] P. Daumar, C. A. Wanger-Baumann, N. V. K. Pillarsetty, L. Fabrizio, S. D.
Carlin, O. A. Andreev, Y. K. Reshetnyak, J. S. Lewis, Bioconjugate Chem.
2012, 23, 1557–1566.
[3] C. Wꢄngler, R. Schirrmacher, P. Bartenstein, B. Wꢄngler, Curr. Med. Chem.
2010, 17, 1092–1116.
[4] D. Thonon, C. Kech, J. Paris, C. Lemaire, A. Luxen, Bioconjugate Chem.
2009, 20, 817–823.
[5] J.-H. Chun, V. W. Pike, Eur. J. Org. Chem. 2012, 4541–4547.
[6] K. Bruus-Jensen, T. Poethko, M. Schottelius, A. Hauser, M. Schwaiger,
H. J. Wester, Nucl. Med. Biol. 2006, 33, 173–183.
[7] S. Abad, P. Nolis, J. D. Gispert, J. Spengler, F. Albericio, S. Rojas, J. R. Her-
ance, Chem. Commun. 2012, 48, 6118–6120.
[32] N. Gorinski, N. Kowalsman, U. Renner, A. Wirth, M. T. Reinartz, R. Seifert,
A. Zeug, E. Ponimaskin, M. Y. Niv, Mol. Pharmacol. Fast Forward 2012,
DOI: 10.1124/mol.112.079137.
[8] M. Namavari, Z. Cheng, R. Zhang, A. De, J. Levi, J. K. Hoerner, S. S. Ya-
ghoubi, F. A. Syud, S. S. Gambhir, Bioconjugate Chem. 2009, 20, 432–
436.
[9] R. Schirrmacher, C. Wꢄngler, E. Schirrmacher, Mini-Rev. Org. Chem. 2007,
4, 317–329.
Received: November 11, 2013
[10] R. Schirrmacher, G. Bradtmçller, E. Schirrmacher, O. Thews, J. Tillmanns,
T. Siessmeier, H. G. Buchholz, P. Bartenstein, B. Waengler, C. M. Niemeyer,
Published online on December 27, 2013
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 337 – 349 349