Towards synthetic adrenaline receptors - Shape-selective adrenaline recognition in water

Article abstract of DOI:10.1002/1521-3765(20020315)8:6<1485::AID-CHEM1485>3.0.CO;2-H

In spite of their key role in signal transduction, the mechanism of action of adrenergic receptors is still poorly understood. We have imitated the postulated binding pattern of the large membrane protein with a small, rationally designed synthetic host molecule. Experimental evidence is presented for the simultaneous operation of electrostatic attraction, hydrogen bonds, π stacking, and hydrophobic interactions. By virtue of this combination of weak attractive forces, adrenaline derivatives in water are bound with high shape selectivity for the slim dopamine skeleton. We think that these findings support the postulated cooperative interplay of noncovalent interactions in the natural receptors. In addition, they provide access to a new type of adrenaline sensor. This may be the first step towards an artificial signal-transduction system.

Full text of DOI:10.1002/1521-3765(20020315)8:6<1485::AID-CHEM1485>3.0.CO;2-H

FULL PAPER
Towards Synthetic Adrenaline Receptors–Shape-Selective
Adrenaline Recognition in Water**
Michael Herm,[b] Oliver Molt,[a] and Thomas Schrader*[a]
Dedicated to Professor Reinhard W. Hoffmann on the occasion of his 68th birthday
Abstract: In spite of their key role in
signal transduction, the mechanism of
action of adrenergic receptors is still
poorly understood. We have imitated
the postulated binding pattern of the
large membrane protein with a small,
rationally designed synthetic host mole-
cule. Experimental evidence is present-
ed for the simultaneous operation of
electrostatic attraction, hydrogen bonds,
p stacking, and hydrophobic interac-
tions. By virtue of this combination of
weak attractive forces, adrenaline deriv-
atives in water are bound with high
shape selectivity for the slim dopamine
skeleton. We think that these findings
support the postulated cooperative in-
terplay of noncovalent interactions in
the natural receptors. In addition, they
provide access to a new type of adrena-
line sensor. This may be the first step
towards an artificial signal-transduction
system.
Keywords: dopamines
rine hormones
recognition ¥ receptors
¥
epineph-
molecular
¥
¥
medical treatment of chronic heart failure has undergone a
remarkable transition in the past ten years. The approach has
changed to a more long-term, reparative strategy, which relies
on treatment with b-adrenergic blocking agents. b-Blockers
have, in fact, become the most extensively studied class of
agents in the treatment of congestive heart failure.[5] Despite
recent advances in the crystallization of membrane-bound
proteins,[6] no crystal structure of any adrenergic receptor has
been resolved to date. Homology modeling of similar
G-protein coupled receptors and site-directed mutagenesis
experiments have provided quite detailed pictures of tertiary
structures, but they still remain somewhat speculative. Thus,
the exact mechanism of adrenaline and antagonist recognition
remains unknown. In addition, the signal transduction path
from the initial adrenaline binding across the membrane into
the interior of the cell resulting in G-protein activation has not
been fully elucidated to date. We think that by synthesizing
small model receptors, chemists can learn from nature about
the efficient interplay of noncovalent interactions, necessary
for efficiency and selectivity in molecular recognition. Such a
small model receptor could allow a systematic study of the
influence of certain noncovalent interactions on the overall
binding enthalpy. It could also shed new light on the specific
combination of noncovalent interactions present in natural
receptors. In our case, the design of an efficient adrenaline
sensor would also open the path for the design of an artificial
signal transduction system.
Introduction
G-protein-coupled receptors (GPCRs) are probably the most
intensively investigated among all receptor groups. 60% of all
commercially available drugs interact with GPCRs, with a
world market volume of 84 billion USD in 1995. Adrenergic
receptors make up one of the major classes in the GPCR
family.[1] They are expressed throughout the human body, and
are involved in vital signal transduction processes from the
extracellular environment across the cell membrane into the
cytosol. For medical treatment, the adrenergic receptor
antagonists are most interesting, because they can suppress
pathological symptoms by blocking the binding sites of the
respective receptors. A well-known example is propranolol,
which is a b-blocker and treats hypertension.[2] b-Adrenergic
receptors are important targets for therapeutic agonists and
antagonists in treatment of heart failure or asthma.[3] The use
of selective a-1A-adrenoceptor antagonists is an efficacious
way to treat benign prostatic hyperplasia.[4] Today adrenergic
receptors are still hot topics for pharmaceutical research. The
[a] Prof. T. Schrader, O. Molt
Fachbereich Chemie der Universit‰t
Hans-Meerwein-Strasse, 35032 Marburg (Germany)
Fax : (‡49)6421-28-25544
E-mail: schradet@mailer.uni-marburg.de
[b] M. Herm
Dupont, Wuppertal (Germany)
In the past decade, the flourishing development of supra-
molecular chemistry combined with the pharmaceutical
interest in catecholamines inspired many groups to design
artificial host molecules for this important class of com-
[**] Towards Synthetic Adrenaline Receptors, Part V. For Part IV, see:
ref. [29].
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author.
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1485 $ 17.50+.50/0
1485

T. Schrader et al.
FULL PAPER
pounds. However, most of these structures were monotopic,
that is, they recognized only one typical functional group of
their guest. Thus, the ammonium group was bound by crown
ethers[7] or ester crowns.[8] Certain cyclopeptides are able to
distinguish between the enantiomeric forms of noradrena-
line.[9] Zinc porphyrin tweezers[10] and xylylene bisphospho-
nates[11] offer the advantage of selectivity for amino alcohols.
Most of these artificial receptors have been developed
primarily for dopamine recognition, and in some cases
dopamine selectivity has been found: a pyrazole-containing
podand imitates the crown ether environment around the
guest;[12] a homooxacalix[3]arene triether was incorporated
into liquid membranes with a PVC matrix, which then showed
a dopamine-selective change in their membrane potential,[13]
and template recognition was achieved with dopamine in
organic inorganic hybrid films prepared by a sol gel pro-
cess.[14] Monotopic catechol recognition was achieved either
with an aza crown, which forms hydrogen bonds to the
catechol hydroxyls,[9] or with bipyridinium moieties, which
exert a p-stacking attraction on the electron-rich catechol
ring.[15] Combined with a peptide tether this has been
exploited for enantioselective dopamine recognition.[16] Some
of these simple binding motifs have been combined with
powerful analytical methods: crown ethers have found
applications in capillary zone electrophoresis,[17] phenylboro-
nates are used for electrochemical detection of catechol-
amines,[18] and w-mercapto poly(ethylene glycol) SAMs (self-
assembled monolayers) on gold electrodes can quantify the
dopamine content in the blood serum.[19] Interdigitated array
microelectrodes have also been used as electrochemical
sensors for catecholamines.[20] In recent years, several ditopic
receptor structures have been developed, which again focus
on ammonium and catechol binding for efficient dopamine
recognition: (aza)crowns bind the ammonium functionality by
means of hydrogen bonds.[21] For catechol recognition a whole
arsenal of different binding motifs has been developed,
ranging from protonated azacrowns for hydrogen bonds with
the phenolic oxygens[22] to macrotricyclic hydrophobic cavities
that include the anionic catechol ring at higher pH values.[23]
hosts.[26] Finally, a bioorganic approach generates dopamine-
selective RNA aptamers by in vitro selection.[27] However,
none of the above-mentioned synthetic receptor molecules is
specific for catechol amino alcohols, and most are far from a
biomimetic recognition pattern.
Our own previous work began with the discovery that in
highly polar organic solution m- and p-xylylene bisphospho-
nates bind to 1,2- and 1,3-amino alcohols one order of
magnitude tighter than to ordinary primary and secondary
amines.[5a] Attachment of aromatic arms on the remaining
phosphonate ester functionality led to second-generation
receptor molecules, which showed a modest increase in
binding energy with catecholamines as a result of p
p
interactions.[5b] The next step was the imitation of the deep
aromatic cleft in the natural adrenoceptor by a hydrophobic
macrocycle with peripheral phosphonates for an induced-fit
process (Figure 1).[28] In 1, the catechol ring is indeed buried in
the macrocyclic hydrophobic cavity. Unfortunately, this new
host undergoes strong self-association in water and cannot
distinguish between amino acids and adrenaline derivatives.
Results and Discussion
To make the fourth generation of biomimetic artificial
adrenaline receptor molecules we followed a new concept:
in order to maximize van der Waals interactions and hydro-
phobic forces, we developed a macrocyclic system with
integral phosphonate moieties (Scheme 1). Their incorpora-
tion into the macrocyclic framework should favor a complex
geometry in which all binding sites of the receptor surround
the adrenaline molecule. Thus, close contacts lead to strong
electrostatic attraction and ideal hydrogen bonds, and also
help desolvation in water. Additional hydrogen bonding sites
for the catechol hydroxy groups and a potential sandwich-type
arrangement for catechol recognition by p stacking were
further features of this new host design. The result was
recently published as the first shape-selective adrenaline host
molecule, which mimics the natural receptor and binds
adrenaline derivatives in water.[29]
p
p Interactions with quinones,[23] hydrogen bonds to
phosphonate anions[24] as well as kinetically fast, reversible
covalent bonds to boronic acids[25] are alternatives for efficient
catechol recognition. Even nonpolar cavities supported by
peripheral carboxylates for nonspecific Coulombic attraction
have been used for the construction of ditopic dopamine
In receptor 2a, the amino alcohol can be bound by the
p-xylylene bisphosphonate moiety, whereas the catechol ring
is flanked by two electron-poor nitroarenes, supported by the
isophthalamide head group for hydrogen bonds to the
phenolic hydroxy groups. This is close to the picture that
Figure 1. Macrocyclic host 1 with a hydrophobic cavity and peripheral phosphonates for ditopic recognition of adrenaline derivatives; left: schematic design;
center: Lewis structure; right: conformational minimum calculated with Cerius2 molecular simulations, force field: Dreiding 2.21.
1486
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1486 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
Scheme 1. Multipoint binding of adrenaline derivatives by biomimetic adrenaline host 2a with integral phosphonates. Left: schematic illustration of the
planned interactions; middle: proposed binding mode; right: natural binding pattern of noradrenaline in the b-adrenergic receptor.
emerged from site-directed mutagenesis studies, molecular
modeling, and electron-diffraction experiments for the natu-
ral example.[30] There, the ammonium functionality is bound
by electrostatic interactions and hydrogen bonds to an
aspartate, enforced by p cation interactions with surround-
ing aromatic amino acid residues. The catechol ring is buried
in a deep cleft between two phenylalanine residues. The
aliphatic as well as both phenolic hydroxy groups are each
hydrogen-bonded to a serine OH (Scheme 1).
In force-field calculations (MacroModel 7.0, Amber*) min-
imum energy structures were found, which correspond to the
postulated arrangement in Scheme 1. High binding enthalpies
result from the combination of electrostatic interactions,
hydrogen bonds, and van der Waals attraction. We carried out
Figure 2. Left: Optimized complex geometry according to Monte Carlo
simulations in water for the inclusion of noradrenaline inside the cavity of
macrocyclic host 2a. Right: Superimposed snapshots of the subsequent
molecular dynamics calculation.
Monte Carlo simulations in water (3000 steps), followed by a
molecular dynamics run at ambient temperature for 10 ps.
The resulting optimized complex geometries are depicted in
Figure 2. In the complex, noradrenaline fits snugly into the
cavity formed by the macrocyclic host. However, a relatively
high degree of flexibility is still maintained, as demonstrated
by the stacked plot of 10 snapshots from the molecular
dynamics calculations. Whitesides and others have shown that
benzylic bonds especially contain a lot of torsional entropy,
resulting from the unhindered rotation around these bonds.[31]
Our macrocycle possesses eight benzylic bonds, which might
reduce the degree of preorientation, but on the other hand
facilitate an induced-fit process.
to the bisphosphonate, and finally close the macrocyclic ring
by a double amidation with isophthaloyl dichloride.
The main synthetic challenge of the whole pathway resides
in the highly and asymmetrically functionalized diphenyl-
methane derivative. This unit has been a favorite structural
key element in artificial receptor molecules for years.[32] It
combines a relatively high degree of preorganization (owing
to the absence of multiple torsional degrees of freedom) with
the potential to create an electron-rich environment in its
concave inner sphere, fine-tuned by appropriate substituents
Synthesis: The encouraging
modeling results prompted us
to develop a modular, highly
convergent synthesis for 2a,
which allows for the simple
construction of structural ana-
logues. Two synthetic cuts lead
to three smaller building blocks,
namely an activated p-xylylene
bisphosphonate, a functional-
ized diphenylmethane and
an
isophthaloyl
derivative
(Scheme 2). We decided to be-
gin with the diphenylmethane
centerpiece, attach two of them
Scheme 2. Retrosynthetic analysis of macrocyclic host 2a: three building blocks open the path to a flexible
modular synthesis with many variations.
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1487 $ 17.50+.50/0
1487

T. Schrader et al.
FULL PAPER
on its phenyl rings.[33] However, most of the open-chain or
macrocyclic hosts based on this moiety were restricted to
symmetric geometries because the general route to diphenyl-
methane derivatives employs the double attack of form-
aldehyde on two electron-rich arenes.[34] Consequently, most
of the published synthetic receptor molecules that carry a
diphenylmethane unit show only small selectivity for unsym-
metrical or even chiral substrates. To improve their recog-
nition pattern, the regioselective introduction of specific
binding sites on the diphenylmethane skeleton would be
beneficial. Few synthetic pathways exist that lead to unsym-
metrical diphenylmethanes: one of them is an acid-catalyzed
Friedel Crafts alkylation of an electron-poor benzyl alcohol
with another arene.[35] However, many benzyl alcohols
polymerize under strongly acidic conditions, and the directing
influence of their substituents may dictate an undesired
chemoselectivity.
Scheme 3. First experiments in coupling of simple molecules to yield
unsymmetrical diphenylmethanes for use in adrenaline receptor synthesis.
agent with the simple nitro-substituted bromobenzene pro-
ceeded smoothly; even the introduction of an additional
phthalimidomethyl group in the electrophilic bromobenzene
does not disturb the coupling reaction (5a/b, Scheme 3).
However, the yield dropped drastically when, instead of
simple benzyl bromide, the TBS-protected phenol was used
(5c, Scheme 4). The problem resided in the metallation step:
We think that 2 is a good example for a highly demanding
diphenylmethane synthesis, because it carries three acid-,
base-, or nucleophile-sensitive groups that have to be
selectively protected and later deprotected for the construc-
tion of the macrocyclic receptor molecule. After trying the
classical methods described above without success, we turned
to organometallic chemistry. In principle, benzylic halides can
be alkylated by a variety of metallated arenes, such as
organolithium, cuprate, or Grignard reagents.[36] Some of
these have already been used to furnish unsymmetrical
diphenylmethanes with electron-rich arenes.[37] We tried all
of these, combining O-THP(tetrahyropyranyl)-protected nu-
cleophiles with simple nitrobenzyl halides. Starting from
4-amino-2-nitrotoluene, we prepared the phthalimide-pro-
tected as well as the butyloxycarbonyl(Boc)-protected benzyl-
amines 3c and 3e with a reactive halide in the para position
(see Scheme 5). The other part came from m-cresol, which
was converted into the corresponding O-THP- or O-TBS-
(tert-butyldimethylsilyl)-protected benzyl bromides. Howev-
er, in all cases, the highly reactive organometallic reagent
attacked the nitro group. Organozinc compounds are much
milder,[38] but for an efficient cross-coupling reaction they
must be activated, for example by conversion into zinc
alkylcuprates.[39] However, even these can engage in side
reactions with nitroarenes. An interesting alternative is the
Pd0- or Ni0-catalyzed cross-coupling of benzylzinc reagents
with bromo- or iodoarenes, introduced by Negishi.[40] This
very mild and highly efficient procedure operates at ambient
temperature and consists of two steps: first a benzyl bromide
is treated in THF with metallic zinc to generate the benzylzinc
reagent without homocoupling, then the resulting solution is
slowly added to a mixture of aryl bromide or iodide,
bis(triphenylphosphanyl)palladium dichloride, and diisobutyl-
aluminum hydride (DIBAL-H) in the same solvent. With
rigorous exclusion of air and humidity, the reaction is
complete after several hours and typically affords products
with 70 95% yield (Scheme 3).
Scheme 4. More simple molecule couplings yielding unsymmetrical
diphenylmethanes.
it came from the unreactive benzyl bromide, which needed
elevated temperatures to form the corresponding benzylzinc
reagent, with the consequence of the unwanted homocoupling
side reaction. The low yield (around 10 30%) was only
marginally raised to ꢀ40%, when instead of the bromoben-
zene the iodo analogue was used. Although this showed that a
more reactive aromatic halide produced a more reactive Pd
reagent, the solution to the problem of unsatisfactory yields
had to be sought in a more reactive benzyl bromide.[41]
Replacement of the TBS protecting group with an acetyl
one finally led to a reproducible 60% total yield in the Negishi
coupling step (5d, Scheme 5).
For solubility reasons and for more convenient deprotec-
tion, the phthaloyl group (3c) was replaced with a Boc moiety
(3e). The new, acidic amide NH functionality did not interfere
with the attack of the benzylzinc reagent on the Pd
intermediate; again, clean conversion of both starting materi-
als was observed. The orthogonal protection strategy allows
selective removal of both protecting groups; this may become
very important if it is not clear at the beginning of the
synthesis at which position the final macrocyclization should
take place. In our case, it is possible either to liberate only the
phenol with potassium carbonate in absolute methanol, or to
deprotect selectively the amine with 50% trifluoroacetic acid
(TFA) in dichloromethane, both at room temperature.
We believe that this example demonstrates the versatility of
the Negishi coupling reaction, which is an extremely mild
reaction and therefore tolerates a wide variety of acid and
base-sensitive groups, including the notorious disturber of
We started our coupling attempts with simple precursors
and systematically included more of the functional groups in
2, in order to explore the scope and limitations of this method
for the construction of various unsymmetrical diphenyl-
methanes. Reaction of the unfunctionalized benzylzinc re-
1488
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1488 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
of 6 with 5e gave the U-type
predecessor 7a, which was Boc-
deprotected under mild condi-
tions with dry trifluoroacetic
acid to give 7b (Scheme 5). The
critical macrocylization step was
subsequently carried out under
high-dilution conditions with a
motor-driven precision pump.
The diamine 7b was thus cy-
clized with isophthaloyl dichlor-
ide, leading to the macrocyclic
bisphosphonate 8a in 38% yield.
Benzene was added as a tem-
plate that has been shown to
bring both nitroarenes into close
proximity by way of a sandwich
arrangement in previous macro-
cyclizations,[42] and thus facili-
tates the double amide connec-
tion with the dicarboxylic acid
dichloride. In the final step, lith-
ium bromide was used as a mild
nucleophile, which selectively
cleaved both methyl esters on
the bisphosphonate, leaving the
aryl esters intact.[43] Macrocyclic
host 2a was obtained in 1.3%
overall yield (12 steps) as a col-
orless hygroscopic solid, soluble
in a wide range of polar solvents
ranging from DMSO to water.
We would like to point out that
ordinary alkyl and aryl phospho-
nates are more sensitive towards
acid and base hydrolysis than
their carboxylate counterparts.
However, they tolerate both hy-
drazine hydrate and dry tri-
fluoroacetic acid at room tem-
perature, so that phthalimide-
and Boc-protected amines can
be selectively deprotected in
their presence.
Scheme 5. Modular and convergent synthesis of hosts 2a and 2b from 4-amino-2-nitrotoluene, m-cresol, and p-
xylylene bisphosphonic acid dimethyl ester dichloride; a) 1. NaNO2, 2. KI (61%); b) NBS, CCl4 (46%); c) K
phthalimide, [18]crown-6, toluene (94%); d) N2H4, ethanol (65%); e) Boc2O, CH2Cl2 (95%); f) Ac2O (89%);
g) NBS, CCl4 (47%); h) 1. Zn, reflux, 2. [Pd(PPh3)4]/DIBAL-H (58%); i) K2CO3, methanol, RT (94%);
j) CH2Cl2, Et3N (57%); k) TFA, CH2Cl2, 08C (99%); l) isophthaloyl chloride or pyridine-2,6-dicarboxylic acid
dichloride, Et3N, THF, benzene, RT (8a: 38%; 8b: 32%); m) LiBr, acetonitrile, 808C (2a: 66%; 2b: 82%).
organometallic reactions, the nitro group. However, for
retrosynthetic analyses one should always bear in mind that
the aromatic precursor for the benzylzinc reagent must be as
electron-poor as possible. Slightly electron-donating (‡I or
‡M effect) groups can severely slow down or completely
prevent the formation of this organozinc intermediate. This
problem, however, can often be circumvented by the choice of
an alternative protecting group, as demonstrated above.
After selective deprotection of the O-acetyl group in 5d
(Scheme 5), a reagent had to be found which allows mono-
activation of a bisphosphonate at both ends. In our hands,
partial hydrolysis of p-xylylene tetramethyl bisphosphonate to
the bismonoester, followed by reaction with oxalyl chloride in
DMF at room temperature, proceeded smoothly and furnish-
ed the bis(esterchloride) 6 in high yields. Double esterification
The above-described modular general approach to our new
adrenaline hosts allows for the convenient preparation of
analogues. Thus, we replaced the isophthalamidic head group
by a pyridine-2,6-dicarboxamidic unit for a more efficient
recognition of the catechol hydroxyls. This could be done in
the penultimate step of the whole synthesis, by simply
carrying out the macrocyclization with pyridine-2,6-dicarbox-
ylic acid dichloride. By this simple variation, the pyridine-
containing host precursor 8b was obtained in 32% yield
(overall yield 1.3%, 12 steps; Scheme 5).
Figure 3 depicts the problem with the isophthalamide head
group: In 2a, both amide hydrogens are involved in repulsive
interactions with the aromatic ortho-proton. This effect can be
clearly seen in molecular mechanics calculations irrespective
of the chosen force-field (Figure 3, top). In consequence, one
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1489 $ 17.50+.50/0
1489

T. Schrader et al.
FULL PAPER
Figure 3. Top: Repulsive NH CH interactions in isophthalamides; two views of the twisted head group in 2a. Bottom: Preorganization in pyridine-2,6-
carboxamides; flat head group in 2b.
of the trans-amide groups could twist, leading to a distorted
overall conformation of the whole macrocycle. In this
conformation, the amide carbonyl could even form a weak
hydrogen bond with the aromatic ortho-proton. Such behav-
ior is often observed with isophthalamides.[44] It can be
circumvented by replacing the isophthalamide by a pyridine-
2,6-dicarboxamide as described above. Here, the pyridine
nitrogen atom forms two intramolecular hydrogen bonds with
the two amidic hydrogens, so that the top part of the host
molecule should be ideally preorganized (2b, Figure 3,
bottom).[45] In the twisted conformation, the amide carbonyl
is now interfering with the basic pyridine nitrogen atom,
leading to repulsion of their respective lone pairs. Vˆgtle et al.
recently presented a molecular knot, whose unique structure
was stabilized by such pyridine-2,6-dicarboxamides.[46]
Initial binding experiments: Initial NMR-spectroscopic bind-
ing studies of 2a with various adrenaline-type guests showed
that 1:1 complexes were formed in all cases (Job plots in
various solvents; see, for example, Figure 4, top).[47]
An ESI spectrum from an equimolar mixture of noradrena-
line hydrochloride and 2a (10À7 m in methanol) produced a
Figure 4. Top: Job plot for complex formation between host 2a and
noradrenaline hydrochloride (CHN proton) in D2O/methanol (1:1);
bottom: NMR titration curve showing the complexation-induced shifts
(CIS; CHN and CHO protons) for complex formation between host 2 and
noradrenaline hydrochloride in methanol.
clean molecular ion peak for the 1:1 complex at m/z 1047, but
no peaks for higher oligomers (Figure 5).
From FT-IR experiments we obtained strong indications for
À
the postulated hydrogen bonds: both P O valence bands of
host 2a were shifted towards smaller wavenumbers in the
complex with noradrenaline. This is in accord with the strong
ion-pair reinforced hydrogen bond between the phosphonate
groups and the guest×s ammonium functionality. In addition,
the amide carbonyl band in 2a was also shifted towards lower
wavenumbers, corresponding to the related hydrogen bond
between the NH group and the catechol oxygen atom. In
order to gain more information about the complex geometry,
we performed NOESY experiments in DMSO, where even
hydrogen bonds with the catechol hydroxyls might be
detectable (Figure 6).
between protons 2 and 1, accompanied with a medium NOE
for proton 2 with 3; together with the absence of any NOE
between protons 2 and 6, this shows that the phosphonate
moieties are pointing inwards into the cavity of macrocycle
2a, with some deviation from the ideal 908 angle between both
arenes. In DMSO, it is conceivable that the lithium counter-
ions form a chelate bridge between both phosphonate anions,
leading to a strong preorientation favorable for the inclusion
of ammonium guest molecules.
The combination of NOEs from protons 7 to their
neighbors is also quite intriguing: a close contact is made to
proton 3, there are medium NOEs to protons 8 and 10, but no
NOE can be observed to proton 4. This indicates that the
methylene protons of the diphenylmethane centerpiece are
In the free host molecule, strong intramolecular NOEs
reveal several steric relationships of key protons essential for
the overall conformation of 2a. A strong NOE is found
1490
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1490 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
Figure 5. ESI-MS for the 1:1 complex of 2a with noradrenaline hydrochloride (mass range m/z 800 1200). Samples (20 mL) were introduced as 10À7
m
‡
‡
‡
solutions in methanol at flow rates of 20 mLminÀ1. The major peaks are: m/z 876: 2aÀ, 873: 2a2À ‡ Li , 890: 2a3À ‡ 2Li , 1047: 2a2À ‡ noradrenaline , 1054:
‡
‡
2a3À ‡ noradrenaline ‡ Li .
were very pleased to find several distinct intermolecular
NOEs between 2a and its guest. Their synopsis in Figure 7
demonstrates that noradrenaline is situated inside the cavity
of 2a, with the amino alcohol in the region of the bisphos-
phonates and the catechol close to the nitroarenes, reaching
up to the isophthalamide head group.[48]
Figure 6. Proton assignments for intramolecular NOE measurements.
Left: macrocycle 2a; right: noradrenaline. Bold arrows: strong NOEs;
thin arrows: medium NOEs.
Figure 7. Intermolecular NOEs in the complex between host 2a and
noradrenaline. The carbon-bound hydrogen atoms have been omitted for
clarity. Bold characters show IR-sensitive functional groups involved in
hydrogen bonds.
always endocyclic, with the nitroarenes at an average angle of
908 to the phenol esters. Again this is exactly the conformation
necessary for inclusion of adrenaline-type guests.
The existence of strong NOEs between protons 13 and 12,
but medium NOEs between 12 and 14, confirms that the
isophthalamide head group preferentially adopts the intended
conformation, with some twist or flexibility of the benzene
ring. This should be improved by the additional intramolec-
ular hydrogen bonds in the pyridine-2,6-dicarboxamide.
If the NOE pattern in the complex of noradrenaline with
macrocyclic host 2a is compared to that of the free binding
partners, the most striking observation is the fact that there is
no significant change at all. Noradrenaline is bound in its
thermodynamically favorable, bioactive conformation, by a
host molecule that barely alters its geometry and shape. We
It is noteworthy that on complexation with 2a or 2b the two
former shift-isochronic hydroxy protons of noradrenaline split
and one of them is drastically shifted to lower field by 1.2 ppm,
whereas the other one is even shifted upfield by 0.3 ppm.
Evidently, the equilibrium of the intramolecular OH ¥¥¥ O
hydrogen bonds in the catechol is strongly shifted towards one
side by complex formation with the host molecule. This can be
explained by the postulated hydrogen bond between the
isophthalamide NH groups and the catechol×s p-oxygen atom
(Figures 8 and 9). The above-described NOESY experiment
confirms this picture: a distinct NOE is observed between
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1491 $ 17.50+.50/0
1491

T. Schrader et al.
FULL PAPER
bisphosphonate moiety. Like 1, macrocyle 2a bears some
similarity to phospholipids, although its amphiphilic nature is
less pronounced.
Selectivity: In order to quantify the contribution of specific
noncovalent interactions to the overall free binding enthalpy
DG, we systematically truncated the guest structure, starting
from adrenaline (Scheme 6; Table 1).
Figure 8. Shift in the catechol hydrogen bond equilibrium by complexation
with host 2a and 2b; experimental evidence from chemically induced shifts
(CIS) and NOE measurements.
Deletion of the N-methyl group leads to an improvement in
binding energy of roughly 1 kJ, probably a steric effect. If the
aliphatic hydroxy group is removed, the association constant
rises again a little, so that we must assume that, contrary to the
case in DMSO, in water host 2a does not recognize amino
alcohols. On truncation of both phenolic hydroxy groups,
however, the free binding enthalpy decreases by more than
2 kJ. In combination with the NOESY experiment and the
downfield shift of one of the phenolic hydroxy protons, we
now have strong experimental evidence for catechol recog-
nition by the isophthalamide head group. Although upfield
shifts occurred in the benzene protons of host and guest, no
shift of the extinction maximum could be observed in the UV
spectra of the complex. Presumably the high flexibility of this
host region prevents formation of discrete charge transfer
complexes.
Finally, if the guest×s phenyl ring is also eliminated, another
marked drop in binding energy of 1.5 kJ molÀ1 is the
1
consequence. In the H NMR spectra, several aromatic host
Figure 9. Drifting catechol OH signals during complex formation between
pyridine-containing host 2b and noradrenaline in DMSO (d ˆ 9.85: amide
NH of host 8, increasing during titration). The relative ratios of equivalents
of 2b to noradrenaline from bottom to top are: 0, 0.25, 0.50, 0.75, 1.00, 1.25,
2.00, 4.75.
and guest protons shift upfield by up to almost 0.5 ppm on
complexation. We conclude that these findings support the
postulated p-stacking interactions between the catechol and
the two nitroarenes in 2a. In summary, the binding constant
protons e and f, but no NOE can be detected between protons
h and g.
Table 1. Binding constants of complexes of host 2a and various guest molecules
from NMR titrations in D2O/MeOD ˆ 1:1.
We performed NMR titrations of macrocyclic host 2a with
noradrenaline in various polar solvents and calculated asso-
ciation constants from the binding curves with nonlinear
regression methods.[49] In DMSO the binding constant is
ꢁ10000mÀ1, in methanol ꢁ1000mÀ1, and in methanol/water
(1:1) ꢁ220mÀ1. This 50-fold drop from DMSO to 50%
methanolic solution is much smaller than in the case of the
open-chain bisphosphonates (ꢁ5000-fold), where the molec-
ular recognition of adrenaline derivatives relies almost
exclusively on electrostatic interactions and hydrogen bonds.
Host 2a must exert additional attractive forces on the guest,
which operate especially effectively in water. In pure water,
host 2a undergoes a self-asso-
Com- Guest molecules[a] Ka(1:1)
DG
[kJmolÀ1
Ddsat
[ppm][c]
Stoichi-
ometry[d]
[b]
[b]
pound
[mÀ1
]
]
9
10
11
12
13
14
15
adrenaline
noradrenaline
dopamine
153 Æ 14% 12.5
215 Æ 12% 13.3
246 Æ 38% 13.6
0.17 Æ 10% 1:1
0.12 Æ 8% 1:1
0.20 Æ 26% 1:1
0.41 Æ 11% 1:1
0.07 Æ 34% 1:1
0.23 Æ 3% 1:1
0.36 Æ 6% 1:1
2-phenylethylamine 102 Æ 14% 11.5
ethanolamine
propranolol
ANP
54 Æ 45% 9.9
204 Æ 5% 13.2
137 Æ 7% 12.1
[a] As hydrochloride salts. [b] Errors are calculated as standard deviations from
the nonlinear regression. [c] Bound shift at 100% complexation, obtained from
the fit (selected CH protons). [d] From Job plots and curve-fitting of the
titration curves.
ciation process, which was de-
termined to be moderately
strong, with a self-association
constant of 270mÀ1. Large
chemically induced shifts (CIS)
with saturation values of up to
almost 1 ppm were found for all
protons in the upper, hydro-
phobic region of 2a (isophthal-
amide and diphenylmethane),
but
they
were
minute
Scheme 6. Guest molecules 9 15 for binding experiments with 2a. The structure of adrenaline has been
systematically truncated to establish the contribution of specific noncovalent interactions.
(<0.2 ppm) in the p-xylylene
1492
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1492 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
for noradrenaline is four times higher than that for simple
ethanolamine, because of a combination of p-stacking and
hydrophobic forces and additional hydrogen bonds with
properly placed recognition sites in the host molecule.
When we tried to include amino acid esters inside the cavity
of 2a, we were surprised: in all cases, regardless of the size of
the ester substituent, there were almost no chemically induced
1
shifts in the H NMR spectrum (Scheme 7). We tried simple
alanine methyl ester, but also aromatic amino acids with a
large electron-rich p surface such as tryptophan and tyrosine
esters. The maximum observed chemical shifts Dd remain
always below 0.03 ppm; upper limits for Ka are estimated at
<10mÀ1. Molecular mechanics calculations do not suggest
that the additional ester group sterically hinders the inclusion
process. Nevertheless, it seems to be a general rule that guests
with an alkyl or aryl substituent a to the N atom do not bind to
2a; thus, a-methyl-4-nitrobenzylamine also shows no chemi-
cally induced shifts on treatment with host 2a. From
attempted Job plots and curve fitting of the ™titration curves∫,
it appears likely that stoichiometric ratios are complex.
Nonlinear regression treatment of the binding curves gave
no saturation in the fit process. Therefore we assume that host
2a rejects amino acid derivatives. This is in sharp contrast to 1,
which could not distinguish between amino acids and adrena-
line derivatives. Hence, the new adrenaline host 2a is shape-
selective for the slim dopamine skeleton (guests 9 15,
Scheme 6). However, the corresponding binding constants
with guests of this type (ꢁ102 mÀ1) are still three orders of
magnitude away from the natural example (ꢁ105 mÀ1).
Figure 10. Top: Job plot for complex formation between host 2b and
noradrenaline hydrochloride (CHN proton) in D2O/methanol (1:1);
bottom: NMR titration curve showing the complexation-induced shifts
(CIS, CHN proton) for complex formation between host 2b and
noradrenaline hydrochloride in the same solvent.
locked in the expected conformation. This can be deduced
from the very weak NOE between protons 12 and 14
A new host with a pyridine-2,6-dicarboxamide head group:
With the replacement of the isophthalamide by a pyridinedi-
carboxamide head group in 2b we hoped to improve the host×s
affinity towards adrenaline derivatives by a higher degree of
preorganization. Again, Job plots revealed a clear 1:1
stoichiometry in a 1:1 mixture of water and methanol
(Figure 10, top). In the negative mass range of an ESI
experiment we obtained a clean, albeit small molecular ion
peak for the 1:1 complex between propranolol hydrochloride
(14) and 2b (m/z 1136). The NOESY study of the free host
molecule, however, showed a first deviation from the
behavior of host 2a: several new intramolecular NOEs
appeared in the diphenylmethane moiety, while others could
not be detected, pointing to a different, more flattened
conformation in this critical region. In contrast to the case for
2a, the pyridine dicarboxamide head group is now indeed
3
(compare Figure 6) and the increased J coupling constant
between the amide NH and the neighboring methylene
protons 11. In the complex with noradrenaline, the guest
structure remains in its bioactive conformation, but the host
structure reverses the changes found for the free host. The
complexation process thus forces the bent host 2b into a
conformation similar to that of the free host 2a, which is much
better suited for the inclusion of the guest. This behavior is
often seen in enzymes and natural receptors, and is called
induced fit. Unfortunately, only one intermolecular NOE
could be detected in this complex, pointing to a somewhat
lower binding affinity.
In pure water, the pyridine-containing host 2b self-asso-
ciates much more strongly than 2a, as dilution experiments
demonstrate. Saturation values of chemically induced shifts
reach 1.2 ppm, and the average
self-association constant is cal-
culated at 1200mÀ1. This could
result from the higher degree of
preorganization for the pyridi-
nedicarboxamide NH groups
pointing into the interior of
the macrocycle. With the lone
pair on the pyridine nitrogen,
these are already involved in
intramolecular hydrogen bonds
and are hence less exposed to
the exterior of 2b.
Scheme 7. More guest molecules 16 20 for binding experiments with 2a. All these molecules carry an additional
alkyl or aryl substituent a to the N atom and are rejected by the macrocyclic host molecule. Note that even the
smallest amino acid derivative l-alanine methyl ester (18) does not bind to 2a.
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1493 $ 17.50+.50/0
1493

T. Schrader et al.
FULL PAPER
For NMR titrations of 2b we used essentially the same
guests already examined for 2a (Schemes 6 and 7). Since the
only structural modification was introduced in the nonpolar
head region of the host molecule, only those guests were
chosen that could be expected to reach into this nonpolar
cavity (unlike ethanolamine, for example). In DMSO, nora-
drenaline binds to 2b with a Ka value of ꢁ4500mÀ1, indicating
a weaker association than with 2a. This is further supported
by NMR titrations in a 1:1 mixture of deuterium oxide and
methanol: all binding constants remain some 30% below the
values obtained with 2a (Table 2). Adrenaline is bound
especially weakly: although chemically induced shifts were
quite strong, in several cases the nonlinear regression did not
converge, and if it did, it furnished very low binding constants.
The additional methyl group in the adrenaline guest seems to
hinder the inclusion of this guest into the cavity of 2b much
more seriously than it does with 2a. Noradrenaline and
dopamine, however, afford smooth binding curves with a good
theoretical fit. These two curves do not differ from each other,
giving zero selectivity for amino alcohols. Nevertheless,
deletion of the catechol hydroxyls leads to a marked decrease
in their association constants, which is even more pronounced
than with 2a. Thus, the pyridine-containing macrocycle 2b is
fairly selective for catecholamines in aequeous solution.
Obviously, the improved preorganization of 2b in the area
of the pyridinedicarboxamide head group enhances the
hydrogen bonding with the catechols, whereas the bent
overall conformation in the diphenylmethane center region
is detrimental to binding in general. Guests with an alkyl or
aryl substituent in the a-position are again completely
rejected by host 2b. Very small shift differences and the
complete lack of convergence in any binding isotherm suggest
that, in water, amino acid esters are almost not complexed at
all. We conclude that the pyridine-containing host 2b under-
goes a slight conformational shift, partly blocking the
entrance to the internal cavity and hence producing somewhat
smaller binding constants than 2a. On the other hand, it is
even more selective against a-substituted alkylammonium
ions such as amino acid esters and related compounds.
Figure 11. Left: Optimized complex geometry according to Monte Carlo
simulations in water for the inclusion of noradrenaline inside the cavity of
macrocyclic host 2b. Note the perfect preorganization by the pyridinecar-
boxamide head group, but also the kinked geometry in the diphenyl-
methane moiety. Right: Superimposed snapshots of the subsequent
molecular dynamics calculation. Note the higher conformational flexibility
of this complex compared with the corresponding assembly of noradrena-
line with 2a.
coplanar to each other, and the catechol moiety of noradrena-
line is ideally sandwiched between them. In addition,
numerous hydrogen bonds are formed with all possible
hydrogen bond donors and acceptors available in that region
of the complex. These findings are in full accord with the
above-described NOESY experiments. The pyridine-contain-
ing host 2b, however, adopts a conformation in the complex
which is much more kinked in the diphenylmethane moiety
than that of 2a. Here, the preorganized upper part of the host
shrinks the cavity slightly. Thus, the guest is pushed a little out
of the macrocycle, which together with the induced fit may
explain why binding constants are consistently lower for
complexes of 2b with adrenaline derivatives than those of 2a.
Conclusion
A biomimetic adrenaline host has been developed that
imitates the combination of all those noncovalent interactions
that are postulated in the natural example. It binds adrenaline
derivatives in water/methanol (1:1) with association constants
in the range of 102 mÀ1. Compared to the natural receptor,
which binds adrenaline with a Ka of 105 mÀ1, this is still three
orders of magnitude away. However, it has to be taken into
consideration that the molecular weight of the natural
receptor is also 40 times higher than that of 2a or 2b. Both
It is interesting to compare the results of the Monte Carlo
simulations for the complexes of noradrenaline and 2a or 2b.
Both conformational searches were conducted under identical
conditions with the same starting geometry, obtained from
simple molecular mechanics calculations. In Figure 11, the
preorganizing effect of the pyridinecarboxamide is beautifully
illustrated: both nitroarene sidewalls are held perfectly
Table 2. Binding constants in complexes of host 2b and various guest molecules (with 9 14 lacking and 17 19 carrying an alkyl or aryl substituent a to the N
atom) from NMR titrations in D2O/MeOD ˆ 1:1.
[b]
[b]
Compound
Guest molecules[a]
Ka(1:1) [mÀ1
]
DG [kJmolÀ1
]
Ddsat [ppm][c]
Stoichiometry[d]
9
10
11
12
14
17
19
adrenaline
noradrenaline
dopamine
2-phenylethylamine
propranolol
21 Æ 172%
136 Æ 10%
142 Æ 14%
50 Æ 31%
7.5
12.2
12.3
9.7
0.47 Æ 165%
0.13 Æ 8%
1:1
1:1
1:1
1:1
1:1
complex
complex
0.24 Æ 11%
0.61 Æ 28%
0.47 Æ 13%
no saturation
no saturation
201 Æ 17%
weak binding
weak binding
13.1
l-tyrosine methyl ester
d-tryptophan methyl ester
[a] As hydrochloride salts. [b] Errors are calculated as standard deviations from the nonlinear regression. [c] Complexation-induced shift at 100%
complexation, obtained from the fit (selected CH protons). [d] From Job plots and curve-fitting of the titration curves. [e] Maximum observed chemical shifts
Dd <0.03 ppm; upper limits for Ka are estimated at <10mÀ1
.
1494
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1494 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
(126 MHz, CDCl3, 258C): d ˆ 38.3 (s; 7), 92.1 (s; 1), 123.7 (s; 10), 130.0 (s;
receptor molecules show a pronounced shape selectivity for
the slim dopamine skeleton.
4), 131.3 (s; 9), 131.7 (s; 5), 133.8 (s; 2), 134.5 (s; 11), 142.5 (s; 6), 148.4 (s; 3),
À
À
ˆ
167.7 (s; 8); FT-IR: nÄ ˆ 3098 (CAr H), 3029 (CAr H), 1772 (C O), 1712
In the future, we want to improve both the binding
efficiency and the selectivity by including structural elements
with a higher degree of preorganization. Above all, we will
substitute the multiple benzylic bonds with rigid moieties such
as tolane spacers and use amide or ester bonds. As Dougherty
(C O), 1532 (-NO2), 1335 (-NO2), 1113 (C I) cmÀ1; MS (CI, NH3, 2008C):
ˆ
À
‡
m/z: 426 ([M‡NH4] ); elemental analysis calcd (%) for C15H9N2O4I
(408.15): C 44.14, H 2.22, N 6.86; found: C 44.26, H 2.39, N 6.61.
2-Aminomethyl-5-iodonitrobenzene (3d): Compound 3c (9.12 g,
22.34 mmol) was refluxed in ethanol (110 mL) with hydrazine hydrate
(1.09 mL, 22.34 mmol) for 3 h. Subsequently 1n HCl (45 mL) was added
and the solution was evaporated to dryness. Acetone (100 mL) was added
to the residue, and the resulting suspension was refluxed with vigorous
stirring for 15 min. The white precipitate was filtered off, washed with
acetone (100 mL), and dried. Then it was treated with water (65 mL) and
diethyl ether (15 mL), and 1n NaOH (45 mL) was slowly added dropwise
while the mixture was stirred. The aqueous layer was extracted three times
with diethyl ether, and the combined organic layers were dried over sodium
sulfate and evaporated. A reddish oil was obtained, which quickly turned
brown in air. Yield: 4.05 g (14.57 mmol, 65%); 1H NMR (500 MHz,
CDCl3,): d ˆ 1.95 (brs, 2H; e), 4.05 (s, 2H; d), 7.37 (d, 3J(H,H) ˆ 8.1 Hz,
1H; b), 7.91 (dd, 3J(H,H) ˆ 8.1 Hz, 4J(H,H) ˆ 1.8 Hz, 1H; c), 8.28 (d,
4J(H,H) ˆ 1.8 Hz, 1 H; a); 13C NMR (126 MHz, CDCl3): d ˆ 42.5 (s; 7), 90.2
(s; 1), 131.1 (s; 2), 132.4 (s; 5), 136.9 (s; 4), 141.5 (s; 6), 147.7 (s; 3).
has recently shown,[50]
p cation interactions operate espe-
cially effective in water as opposed to electrostatic interac-
tions. Replacement of the nitroarenes in 2 with pyridinium
moieties may add another stabilizing effect. With these
improved biomimetic adrenaline hosts, we aim at the con-
struction of an artificial signal transduction system operating
across a synthetic membrane.
Experimental Section
4-Iodo-2-nitrotoluene (3a): 4-Amino-2-nitrotoluene (10.0 g, 65.7 mmol)
was suspended in diluted aqueous sulfuric acid (10 vol%, 200 mL) at 08C
and treated slowly with a solution of sodium nitrite (4.76 g, 69.0 mmol) in
water (10 mL). The mixture was stirred for 1 h at 08C. Subsequently it was
filtered into a solution of potassium iodide (15.0 g, 90.4 mmol) and sodium
acetate trihydrate (350 g) in water (300 mL) stirred gently at 08C. Stirring
was continued for 1 h, then the reaction mixture was extracted with diethyl
ether (3 Â 200 mL), twice with 1n aqueous sodium thiosulfate, once with
aqueous ammonia (10 vol%) and finally with water (100 mL each time).
The ethereal layer was dried over magnesium sulfate and evaporated to
dryness. The residue was chromatographed over silica with n-hexane/
dichloromethane (3:2), again evaporated to dryness and recrystallized from
ethanol, furnishing yellow crystals. Yield: 10.58 g (40.2 mmol, 61%); m.p.
2-tert-Butyloxycarbonylaminomethyl-5-iodonitrobenzene (3e): Com-
pound 3d (1.35 g, 4.86 mmol) was dissolved in dry dichloromethane
(16 mL). At 08C triethylamine (0.68 mL, 4.86 mmol) and di-tert-butyl
pyrocarbonate (1.28 mL, 5.58 mmol) were added, and the mixture was
stirred overnight at room temperature. Then the solvent was removed and
the residue was chromatographed over silica gel with dichloromethane
(Rf ˆ 0.14), yielding a pale yellow oil. Yield: 1.75 g (4.63 mmol, 95%);
1H NMR (500 MHz, CDCl3): d ˆ 1.43 (s, 9H; f), 4.50 (d, 3J(H,H) ˆ 6.3 Hz,
2H; d), 5.30 (brs, 1H; e), 7.37 (d, 3J(H,H) ˆ 8.2 Hz, 1H; b), 7.93 (dd,
3J(H,H) ˆ 8.2 Hz, 4J(H,H) ˆ 1.9 Hz, 1H; c), 8.36 (d, 4J(H,H) ˆ 1.9 Hz, 1H;
a); 13C NMR (126 MHz, CDCl3): d ˆ 28.3 (s; 10), 42.1 (s; 7), 80.0 (s; 9), 92.0
(s; 1), 133.1 (s; 2), 133.6 (s; 4), 134.2 (s; 5), 142.8 (s; 6), 148.5 (s; 3), 155.7 (s;
À
À
À
8); FTIR (film): nÄ ˆ 3347 (N H), 3096 (CAr H), 2977, 2933 (C H), 1698
1
3
(C O), 1529 (C O), 1599 (N H), 1529, 1345 (NO2), 1167 (C O) cmÀ1; MS
(CI, NH3, 2008C): m/z: 378 ([M]À); elemental analysis calcd (%) for
C12H15N2O4I (378.17): C 38.11, H 4.00, N 7.41; found: C 38.10, H 4.07, N 7.40.
608C; H NMR (500 MHz, CDCl3): d ˆ 2.54 (s, 3H; d), 7.09 (d, J(H,H) ˆ
ˆ
ˆ
À
À
3
4
8.1 Hz), 1H; b), 7.80 (dd, J(H,H) ˆ 8.1 Hz, J(H,H) ˆ 1.8 Hz, 1H; c), 8.27
(d, 4J(H,H) ˆ 1.8 Hz, 1H; a); 13C NMR (126 MHz, CDCl3): d ˆ 20.1 (s; 7),
89.6 (s; 1), 133.2 (s; 4), 133.2 (s; 2), 134.2 (s; 5), 141.8 (s; 6), 149.6 (s; 3).
General procedure for the Negishi coupling reaction: The reaction had to
be carried out with rigid exclusion of oxygen and humidity. Zinc powder
(0.882 g, 13.50 mmol) and dry tetrahydrofuran (0.5 mL) were treated with
1,2-dibromoethane (0.019 mL, 0.225 mmol). The mixture was heated
carefully with a heat gun until soaplike bubbles appeared and the solution
became slightly clouded. The benzyl bromide reagent (6.75 mmol) was
dissolved in dry tetrahydrofuran (3 mL) and added dropwise at À108C to
the activated zinc (1 h). Subsequently the reaction mixture was stirred for
1 h at ambient temperature. The excess of zinc powder was filtered out
under inert conditions, and a slightly clouded, greenish solution was
obtained. Bis(triphenylphosphanyl)palladium(ii) chloride (0.144 g,
0.206 mmol) was suspended in dry tetrahydrofuran (7 mL) and treated
slowly with a 1m solution of diisobutylaluminum hydride in hexane
(0.411 mL), producing a color change from yellow to green-black. Then a
solution of the aryl halide (4.50 mmol) in dry tetrahydrofuran (10 mL) was
added. Finally the solution of the organozinc reagent was added dropwise,
producing a color change to red-brown, and the reaction mixture was
stirred overnight at ambient temperature. The solvent was removed in
vacuo at a temperature no higher than 508C. The residue was dissolved in
dichloromethane and treated with n-hexane, until a precipitate formed,
which was filtered off. Evaporation to dryness was followed by chromato-
graphic purification over silica gel with dichloromethane/ethyl acetate
(19:1), giving a yellow oil.
2-Bromomethyl-5-iodonitrobenzene (3b): 4-Iodo-2-nitrotoluene (3a,
3.94 g, 14.98 mmol) was dissolved in dry tetrachloromethane (12.5 mL)
and treated with N-bromosuccinimide (NBS; 2.67 g, 14.98 mmol). The
mixture was refluxed; at the beginning and every 4 h a small amount of
dibenzoylperoxide was added. After teh mixture had been cooled to room
temperature, the solid was filtered off and washed with a little dichloro-
methane. The filtrate was evaporated to dryness and chromatographed
over silica gel with dichloromethane. On evaporation of the solvent a solid
was obtained, which was heated for 1 h in boiling petroleum ether 40/60
(30 mL). After cooling to room temperature and filtration, a colorless
powder was obtained. Yield: 2.34 g (6.84 mmol, 46%); m.p. 988C (DSC);
1H NMR (500 MHz, CDCl3): d ˆ 4.76 (s, 2H; d), 7.30 (d, 3J(H,H) ˆ 7.9 Hz,
1H; b), 7.93 (dd, 3J(H,H) ˆ 7.9 Hz, 4J(H,H) ˆ 1.9 Hz, 1H; c), 8.35 (d,
4J(H,H) ˆ 1.9 Hz, 1 H; a); 13C NMR (126 MHz, CDCl3): d ˆ 28.1 (s; 7), 93.5
(s; 1), 132.4 (s; 4), 133.8 (s; 5), 134.1 (s; 2), 142.6 (s; 6), 148.1 (s; 3); FT-IR
À
(KBr): nÄ ˆ 3082 (CAr H), 2857 (-CH2-), 1524 (-NO2), 1344 (-NO2), 1079
(Ar I), 804 (Ar)cmÀ1
;
MS (CI, NH3, 2008C): m/z: 378, 376
À
‡
‡
([M‡NH3‡NH4] ), 361, 359 ([M‡NH4] ); elemental analysis calcd (%)
for C7H5NO2BrI (341.93): C 24.59, H 1.47, N 4.09; found: C 24.62, H 1.77, N
4.27.
2-Phthaloylimidomethyl-5-iodonitrobenzene (3c): Iodonitrobenzene 3b
(8.09 g, 23.66 mmol), potassium phthalimide (5.27 g, 28.47 mmol) and
[18]crown-6 (0.44 g, 2.37 mmol) were stirred for 3 h at 808C in dry toluene
(30 mL) under argon. Subsequently the solution was diluted with dichloro-
methane and insoluble components were filtered off. The filtrate was
evaporated to dryness in vacuo, and the resulting residue was chromato-
graphed over silica gel with dichloromethane/n-hexane (5:1), affording
colorless crystals. Yield: 9.12 g (22.34 mmol, 94%); m. p. 1778C (DSC);
Model diphenylmethane 5a: Yield: 0.68 g (3.01 mmol, 67%); m.p. 358C;
1H NMR: d ˆ 2.57 (s, 3H), 4.02 (s, 2H), 7.18 (d, J ˆ 7.6 Hz, 2H), 7.22 (t, J ˆ
7.6 Hz, 1H), 7.22 (d, J ˆ 8.1 Hz, 1H), 7.30 (dd, J ˆ 8.1 Hz, J ˆ 1.3 Hz, 1H),
7.30 (dd, J ˆ 7.6 Hz, J ˆ 7.6 Hz, 2H), 7.80 (d, J ˆ 1.3 Hz, 1H); 13C{1H} NMR:
d ˆ 20.0, 40.8, 124.8, 126.6, 128.8, 128.9, 131.2, 133.0, 133.5, 139.6, 140.5,
149.3; elemental analysis calcd (%) for C14H13NO2: C 73.99, H 5.77, N 6.16;
found: C 73.75, H 5.87, N 6.21.
3
1H NMR (500 MHz, CDCl3, 258C): d ˆ 5.21 (s, 2H; d), 7.00 (d, J(H,H) ˆ
8.2 Hz, 1H; b), 7.77 (dd, 3J(H,H) ˆ 5.7 Hz, 4J(H,H) ˆ 3.2 Hz, 2H; f), 7.83
(dd, 3J(H,H) ˆ 8.2 Hz, 4J(H,H) ˆ 1.9 Hz, 1H; c), 7.89 (dd, 3J(H,H) ˆ 5.7 Hz,
4J(H,H) ˆ 3.2 Hz, 2H; e), 8.39 (d, 4J(H,H) ˆ 1.9 Hz, 1H, a); 13C NMR
Model diphenylmethane 5b: Yield: 1.07 g (2.88 mmol, 64%); 1H NMR:
d ˆ 4.02 (s, 2H), 5.26 (s, 2H), 7.14 (d, J ˆ 7.6 Hz, 2H), 7.15 (d, J ˆ 8.1 Hz,
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1495 $ 17.50+.50/0
1495

T. Schrader et al.
FULL PAPER
1H), 7.22 (t, J ˆ 7.6 Hz, 1H), 7.29 (dd, J ˆ 7.6 Hz, J ˆ 7.6 Hz, 2H), 7.34 (dd,
J ˆ 8.1 Hz, J ˆ 1.3 Hz, 1H), 7.76 (dd, J ˆ 5.7 Hz, J ˆ 3.2 Hz, 2H), 7.89 (dd,
J ˆ 5.7 Hz, J ˆ 3.2 Hz, 2H), 7.93 (d, J ˆ 1.3 Hz, 1H); 13C{1H} NMR: d ˆ
38.5, 41.0, 123.6, 125.4, 126.7, 128.3, 128.8, 128.9, 129.4, 131.8, 134.1, 134.3,
139.0, 142.3, 148.1, 167.9.
p-Xylylene-a,a'-bis(phosphonic acid monomethyl ester chloride) (6): All
steps had to be carried out with rigorous exclusion of humidity and air. The
above-described bis(monomethylphosphonate) (0.357 g, 1.21 mmol) was
suspended in dry dichloromethane (15 mL) at À108C. Oxalyl chloride
(0.23 mL, 2.67 mmol) and then DMF (2 drops) were slowly added to this
mixture, which was stirred further for 2 h at room temperature and then for
another 2 h at 408C. Complete conversion was indicated by complete
dissolution of the solid. The solvent was condensed off at 408C and
10À2 Torr. The product was extremely sensitive to hydrolysis and was used
directly for the next step without further purification. 1H NMR (500 MHz,
Model diphenylmethane 5c: Yield: 0.65 g (1.35 mmol, ꢁ30%); m.p. 798C;
1H NMR: d ˆ 0.16 (s, 6H), 0.95 (s, 9H), 3.95 (s, 2H), 5.26 (s, 2H), 6.63 (d,
J ˆ 1.9 Hz, 1H), 6.70 (dd, J ˆ 8.2 Hz, J ˆ 1.9 Hz, 1H), 6.71 (d, J ˆ 8.2 Hz,
1H), 7.14 (dd, J ˆ 8.2 Hz, J ˆ 8.2 Hz, 1H), 7.15 (d, J ˆ 8.2 Hz, 1H), 7.33 (dd,
J ˆ 8.2 Hz, J ˆ 1.3 Hz, 1H), 7.77 (dd, J ˆ 5.7 Hz, J ˆ 3.2 Hz, 2H), 7.89 (dd,
J ˆ 5.7 Hz, J ˆ 3.2 Hz, 2H), 7.91 (d, J ˆ 1.3 Hz, 1H); 13C{1H} NMR: d ˆ
À4.4, 18.2, 25.6, 38.5, 40.8, 118.4, 120.8, 121.9, 123.6, 125.4, 128.3, 129.3,
129.7, 131.9, 134.1, 134.3, 140.5, 142.3, 148.1, 156.0, 167.8; elemental analysis
calcd (%) for C28H30N2O5Si: C 66.91, H 6.02, N 5.57; found: C 66.99, H 5.93,
N 5.38.
2
3
CDCl3): d ˆ 3.54 (d, J(H,P) ˆ 19.1 Hz, 4H; b), 3.86 (d, J(H,P) ˆ 13.3 Hz,
6H; c), 7.32 (s, 4H; a); 31P NMR (202 MHz, CDCl3): d ˆ 41.2 (s).
p-Xylylene-a,a'-bis(phosphonic acid mono-3-[3'-nitro-4'-(tert-butyloxycar-
boxyamino-methyl)phenylmethyl]phenylmonomethyl ester (7a): Phospho-
nate ester 6 (0.40 g, 1.21 mmol) was dissolved in dry dichloromethane
(10 mL). Subsequently a solution of 5e (0.87 g, 2.43 mmol) and triethyl-
amine (0.56 mL, 4.01 mmol) in dry dichloromethane (5 mL) was added
quickly by syringe. After 15 min the solution was evaporated to dryness and
the residue was chromatographed over silica gel and eluted with chloro-
form/acetone (5:1). A yellow solid was obtained as a mixture of P
stereoisomers. Yield: 0.68 g (0.70 mmol, 57%); m.p. 2088C (decomp,
DSC); 1H NMR (500 MHz, CDCl3): d ˆ 1.42 (s, 18H; n), 3.29 (d, 2J(H,P) ˆ
20.8 Hz, 4H; b), 3.69 (d, 3J(H,P) ˆ 10.7 Hz, 6H; c), 3.98 (s, 4H; h), 4.52 (d,
3J(H,H) ˆ 6.3 Hz, 4H; l), 5.39 (brs, 2H; m), 6.93 (s, 2H; d), 6.94 (d,
3J(H,H) ˆ 7.6 Hz, 2H; g), 6.98 (d, 3J(H,H) ˆ 8.2 Hz, 2H; e), 7.22 (dd,
3J(H,H) ˆ 8.2 Hz, 3J(H,H) ˆ 7.6 Hz, 2H; f), 7.25 (s, 4H; a), 7.40 (d,
2-tert-Butyloxycarbonylaminomethyl-5-(3'-acetoxyphenylmethyl)-nitro-
benzene (5d): Yield: 1.04 g (2.60 mmol, 58%); 1H NMR (500 MHz, CDCl3,
258C): d ˆ 1.43 (s, 9H), 2.28 (s, 3H), 4.04 (s, 2H), 4.52 (d, 3J(H,H) ˆ 6.3 Hz,
2H), 5.32 (brs, 1H), 6.89 (dd, 4J(H,H) ˆ 1.9 Hz, 4J(H,H) ˆ 1.9 Hz, 1H),
6.98 (dd, 3J(H,H) ˆ 8.2 Hz, 4J(H,H) ˆ 1.9 Hz, 1H), 7.04 (d, 3J(H,H) ˆ
7.6 Hz, 1H), 7.32 (dd, 3J(H,H) ˆ 8.2 Hz, 3J(H,H) ˆ 7.6 Hz, 1H), 7.42 (dd,
3J(H,H) ˆ 8.2 Hz, 4J(H,H) ˆ 1.9 Hz, 1H), 7.54 (d, 3J(H,H) ˆ 8.2 Hz, 1H),
7.87 (s, 1H); 13C NMR (126 MHz, CDCl3, 258C): d ˆ 21.1 (s), 28.3 (s), 40.7
(s), 42.1 (s), 79.8 (s), 120.0 (s), 122.0 (s), 125.2 (s), 126.3 (s), 129.7 (s), 131.8
(s), 132.5 (s), 134.4 (s), 140.9 (s), 141.5 (s), 148.3 (s), 151.0 (s), 155.8 (s), 169.3
À
À
À
(s); IR (film): nÄ ˆ 3367 (N H), 3066 (CAr H), 2978, 2933 (C H), 1765, 1709
3
3J(H,H) ˆ 7.6 Hz, 2H; k), 7.53 (d, J(H,H) ˆ 7.6 Hz, 2H; j), 7.83 (s, 2H; i);
(C O), 1529 (NO2), 1366 (NO2), 1207, 1168 (C O) cmÀ1; MS (CI, NH3,
2008C): m/z: 400 ([M]À); calcd for C21H24N2O6 (400.43): C 62.99, H 6.05, N
6.99; found: C 62.91, H 5.90, N 6.75.
ˆ
À
1
13C NMR (126 MHz, CDCl3): d ˆ 28.3 (s; 21), 32.9 (d, J(C,P) ˆ 139.3 Hz;
2
3), 40.6 (s; 11), 42.1 (s; 18), 53.4 (d (diastereomers), J(C, P) ˆ 3.6 Hz; 4),
79.7 (s; 20), 118.6 (s; 10), 120.9 (s; 6), 125.1 (s; 13), 125.5 (s; 8), 129.6 (d
(diastereomers), 2J(C,P) ˆ 2.4 Hz; 2), 130.1 (s; 9), 130.2 (s; 1), 131.7 (s; 15),
132.6 (s; 16), 134.3 (s; 17), 141.2 (s; 12), 141.5 (s; 7), 148.2 (s; 14), 150.8 (d
(diastereomers), 2J(C,P) ˆ 4.2 Hz; 5), 155.8 (s; 19); 31P NMR (202 MHz,
2-tert-Butyloxycarbonylaminomethyl-5-(3'-hydroxyphenylmethyl)nitro-
benzene (5e): Potassium carbonate (0.36 g, 2.57 mmol) was suspended
under argon in dry methanol (120 mL).
A solution of 5d (1.03 g,
À
À
ˆ
CDCl3): d ˆ 25.5 (s); IR (KBr): nÄ ˆ 3445 (N H), 2977 (C H), 1708 (C O),
2.57 mmol) in dry methanol (25 mL) was slowly added dropwise. After
15 min 1n HCl (5.14 mL) was added dropwise, followed by water (140 mL)
and ethyl acetate as well as aqueous saturated NaCl until phase separation
occurred. The aqueous layer was extracted once more with ethyl acetate,
and the combined organic phases were dried over magnesium sulfate and
evaporated to dryness. The residue was chromatographed over silica gel
with chloroform/acetone (10:1, Rf ˆ 0.48), furnishing a yellow oil. Yield:
0.87 g (2.43 mmol, 94%); 1H NMR (500 MHz, CDCl3): d ˆ 1.42 (s, 9H; l),
3.96 (s, 2H; f), 4.51 (d, 3J(H,H) ˆ 6.3 Hz, 2H, j), 5.35 (brs, 1H; k), 5.39 (brs,
ˆ
À
1607, 1586 (arom), 1530 (NO2), 1366 (NO2), 1249 (P O/P OAr), 1046
(P OMe) cmÀ1; elemental analysis calcd (%) for C48H56N4O14P2 (974.94): C
59.13, H 5.79, N 5.75; found: C 59.12, H 5.85, N 5.65.
À
p-Xylylene-a,a'-bis(phosphonic acid mono-3-[3'-nitro-4'-(aminomethyl)-
phenylmethyl]phenyl monomethyl ester bis(hydrogentrifluoroacetate)
(7b): Ester 7a (0.82 g, 0.841 mmol) was stirred in dry dichloromethane
(35 mL) with trifluoroacetic acid (35 mL) under argon at 08C for 1.5 h.
Afterwards, the solution was evaporated to dryness and treated several
times with dry dichloromethane with subsequent removal of the solvent, in
order to eliminate azeotropically traces of trifluoroacetic acid. The oily
brown crude product was used for the next step without workup. Yield:
0.84 g (0.84 mmol, >99%); 1H NMR (500 MHz, CD3OD): d ˆ 3.47 (d,
2J(H,P) ˆ 20.5 Hz, 4H; b), 3.77 (d, 3J(H,P) ˆ 11.0 Hz, 6H; c), 4.13 (s,
4H; h), 4.41 (s, 4H; l), 6.99 (d, 3J(H,H) ˆ 7.9 Hz, 2H; g), 7.01 (s, 2H; d), 7.12
3
4
1H; a), 6.62 (s, 1H; b), 6.71 (dd, J(H,H) ˆ 8.2 Hz, J(H,H) ˆ 2.5 Hz, 1H;
e), 6.73 (d, 3J(H,H) ˆ 7.6 Hz, 1H; c), 7.17 (dd, 3J(H,H) ˆ 8.2 Hz, 3J(H,H) ˆ
7.6 Hz, 1H; d), 7.41 (dd, 3J(H,H) ˆ 7.6 Hz, 4J(H,H) ˆ 1.3 Hz, 1H; i), 7.51 (d,
3J(H,H) ˆ 8.2 Hz, 1H; h), 7.86 (s, 1H; g); 13C NMR (126 MHz, CDCl3): d ˆ
28.4 (s; 17), 40.9 (s; 7), 42.2 (s; 14), 79.9 (s; 16), 113.7 (s; 6), 115.8 (s; 2),
121.2 (s; 4), 125.2 (s; 9), 130.0 (s; 5), 131.8 (s; 11), 132.2 (s; 12), 134.4 (s; 13),
141.0 (s; 8), 142.1 (s; 3), 148.2 (s; 10), 155.9 (s; 15), 156.1 (s; 1); FT-IR (film):
3
3
(d, J(H,H) ˆ 7.9 Hz, 2H; e), 7.31 (dd, 3J(H,H) ˆ 7.9 Hz, J(H,H) ˆ 7.9 Hz,
2H; f), 7.34 (s, 4H; a), 7.66 (d, 3J(H,H) ˆ 7.9 Hz, 2H; j), 7.70 (dd, 3J(H,H) ˆ
7.9 Hz, 4J(H,H) ˆ 1.6 Hz, 2H; k), 8.11 (d, 4J(H,H) ˆ 1.6 Hz, 2H; i);
À
À
À
À
À
nÄ ˆ 3350 (O H, N H), 3015 (CAr H), 2980 (C H), 2933 (C H), 1689
ˆ
À
À
(C O), 1589 (N H), 1531 (NO2), 1347 (NO2), 1162 (C O), 757
(3CAr H) cmÀ1; MS (CI, NH3, 2008C): m/z: 358 ([M]À); elemental analysis
calcd (%) for C19H22N2O5 (358.39): C 63.68, H 6.19, N 7.82; found: C 63.39,
H 6.47, N 7.78.
13C NMR (126 MHz, CD3OD): d ˆ 33.7 (d, J(C,P) ˆ 138.1 Hz; 3), 41.6 (s;
1
À
11), 42.1 (s; 18), 54.6 (d, 2J(C,P) ˆ 3.6 Hz (diastereomers); 4), 117.8 (q,
1J(C,F) ˆ 284.6 Hz; 19), 120.1 (s; 10), 122.4 (s; 6), 127.1 (s; 15), 127.3 (s; 13),
127.7 (s; 8), 131.4 (d, 2J(C,P) ˆ 3.0 Hz (diastereomers); 2), 131.6 (s; 9), 131.7
(s; 1), 134.7 (s; 16), 136.3 (s; 17), 143.4 (s; 12), 146.4 (s; 7), 150.2 (s; 14),
p-Xylylene-a,a'-bis(phosphonic acid monomethyl ester): p-Xylylene-a,a'-
bis(phosphonic acid dimethyl ester) (4.62 g, 15.52 mmol) was refluxed for
6 h with aqueous NaOH (20vol%, 23 mL). Cooled with ice, the solution
was subsequently treated with half-concentrated HCl (23 mL), giving a
white precipitate. The precipitate was filtered off, washed with water and
acetone and recrystallized from methanol. Yield: 2.95 g (9.16 mmol, 64%);
m.p. 2108C (DSC); 1H NMR (500 MHz, [D6]DMSO): d ˆ 3.04 (d,
2J(H,P) ˆ 20.2 Hz, 4H; b), 3.52 (d, 3J(H,P) ˆ 10.7 Hz, 6H; c), 6.15 (brs,
2H; d), 7.17 (s, 4H; a); 13C NMR (126 MHz, [D6]DMSO, 258C): d ˆ 32.8 (d,
2
2
152.3 (d, J(C,P) ˆ 4.2 Hz (diastereomers); 5), 162.3 (q, J(C,F) ˆ 36.0 Hz;
20); 31P NMR (202 MHz, CD3OD): d ˆ 26.8 (s).
1,10,27,34-Tetraoxo-1,10-dimethoxy-1,10-diphospha-11,50-dioxa-26,35-di-
aza[2](3)benzeno[3](2)benzeno[1](3,4)nitrobenzeno[3](2)benzeno[3](3,5)-
nitrobenzeno[1](2)benzenocyclophane (8a): A solution of 7b (0.843 g,
0.841 mmol) and triethylamine (0.82 mL, 5.89 mmol) in dry THF (70 mL)
and a solution of isophthaloyl chloride (0.171 g, 0.841 mmol) in 70 mL of
dry THF were added simultaneously over 3 h at room temperature evenly
to a mixture from dry THF (60 mL) and dry benzene (15 mL) by syringe.
The reaction mixture was stirred overnight at room temperature, and
subsequently filtered from the precipitated triethylammonium chloride.
The filtrate was evaporated to dryness, and the residue was chromato-
graphed over silica gel with chloroform/acetone (1:1, Rf ˆ 0.33), yielding a
light yellow solid. Yield: 0.29 g (0.321 mmol, 38%); m.p. 2408C (decomp,
DSC); 1H NMR (500 MHz, CDCl3): d ˆ 3.25 (d, 2J(H,P) ˆ 20.8 Hz, 4H; b),
1J(C,P) ˆ 133.2 Hz; 3), 51.7 (d, 2J(C,P) ˆ 3.6 Hz (two signals from
2
diastereomers; 4), 129.8 (s; 1), 131.3 (d, 2J(C,P) ˆ 4.9 Hz (two signals from
2 diastereomers; 2); 31P NMR (202 MHz, [D6]DMSO): d ˆ 26.1 (s); FT-IR
À
À
(KBr): nÄ ˆ 3449 (O H), 2963, 2924, 2858 (C H), 2606 (P(O)OH), 1655
ˆ
À
À
(arom), 1268 (P O), 1138 (C O), 1050 (P OMe), 830 (2 adjacent
CAr H) cmÀ1; MS (CI, NH3, 2008C): m/z: 312 ([M‡NH4] ); elemental
‡
À
analysis calcd (%) for C10H16O6P2 (294.18): C 40.83, H 5.48; found: C 40.50,
H 5.45.
1496
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1496 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
3.67/3.68 (d (diastereomers), 3J(H,P) ˆ 11.0 Hz, 6H; c), 3.96 (s, 4H; h), 4.81
(d, 3J(H,H) ˆ 6.3 Hz, 4H; l), 6.78 (s, 2H; d), 6.91 (d, 3J(H,H) ˆ 8.2 Hz, 2H;
g), 6.97 (d, 3J(H,H) ˆ 7.6 Hz, 2H; e), 7.19 (s, 4H; a), 7.21 (dd, 3J(H,H) ˆ
8.2 Hz, 3J(H,H) ˆ 7.6 Hz, 2H; f), 7.23 (t, 3J(H,H) ˆ 6.3 Hz, 2H; m), 7.37 (dd,
3J(H,H) ˆ 7.9 Hz, 4J(H,H) ˆ 1.3 Hz, 2H; k), 7.48 (t, 3J(H,H) ˆ 7.9 Hz, 1H;
p), 7.60 (d, 3J(H,H) ˆ 8.2 Hz, 2H; j), 7.83 (d, 4J(H,H) ˆ 1.3 Hz, 2H; i), 7.93
(dd, 3J(H,H) ˆ 7.6 Hz, 4J(H,H) ˆ 1.3 Hz, 2H; o), 8.11 (d, 4J(H,H) ˆ 1.3 Hz,
1H, n); 13C NMR (126 MHz, CDCl3): d ˆ 32.8 (d, 1J(C,P) ˆ 139.3 Hz; 3),
40.7 (s; 11), 41.4 (s; 18), 53.4 (d (diastereomers), 2J(C,P) ˆ 3.6 Hz; 4), 118.7
(s; 10), 120.8 (s; 6), 125.2 (s; 13), 125.4 (s; 23), 125.6 (s; 8), 129.1 (s; 21),
129.7 (d (diastereomers), 2J(C,P) ˆ 3.6 Hz; 2), 130.1 (s; 9), 130.2 (s; 1), 130.3
(s; 22), 131.3 (s; 15), 132.6 (s; 16), 134.3 (s; 20), 134.4 (s; 17), 141.1 (s; 12),
142.1 (s; 7), 148.3 (s; 14), 150.7/150.8 (d (diastereomers), 2J(C,P) ˆ 4.2 Hz;
5), 166.4 (s; 19); 31P NMR (202 MHz, CDCl3): d ˆ 25.42/25.43 (s, mixture of
(d, 2J(C,P) ˆ 4.2 Hz (diastereomers); 5), 163.2 (s; 19); 31P NMR (202 MHz,
À
CDCl3, 258C): d ˆ 25.3 (s (diastereomers); FT-IR: nÄ ˆ 3418 (N H), 2957
À
ˆ
À
(C H), 1677 (C O), 1608, 1586 (arene), 1529, 1356 (NO2), 1237 (P OAr),
‡
1043 (P OMe) cmÀ1; MS (FAB ‡ NBA): m/z: 906 ([M‡H] ); elemental
À
analysis calcd (%) for C4H41N5O12P2 (905.79): C 59.67, H 4.56, N 7.73;
found: C 59.40, H 4.81, N 7.59.
1,10,27,34-Tetraoxo-1,10-dioxido-1,10-diphospha-11,50-dioxa-26,35-di-
aza[2](3)benzeno[3](2)benzeno[1](3,4)nitrobenzeno[3](2,1)pyridino[3](3,5)-
nitrobenzeno[1](2)benzenocyclophane dilithium salt (2b): Cyclophane 8b
(144.1 mg, 0.159 mmol) was suspended in dry acetonitrile (1.5 mL) and
treated with a solution of lithium bromide (28.3 mg, 0.326 mmol) in dry
acetonitrile (1.5 mL). The mixture was heated under reflux for one week.
Then the product was filtered off and washed with cold acetonitrile and
diethyl ether. Yield: 116.0 mg (0.130 mmol, 82%); m.p. 2508C (decomp,
DSC); 1H NMR (500 MHz, CD3OD, 258C): d ˆ 2.88 (d, 2J(H,P) ˆ 19.6 Hz,
À
À
ˆ
diastereomers); FT-IR (KBr): nÄ ˆ 3424 (N H), 2956 (C H), 1655 (C O),
ˆ
À
1607, 1585 (arom), 1529, 1348 (NO2), 1260 (P O), 1235 (P OAr), 1044
À
3
4H; b), 3.85 (s, 4H; g), 4.83 (s, 4H; k), 6.63 (s, 2H; c), 6.79 (d, J(H,H) ˆ
‡
(P OMe) cmÀ1; MS (CI, NH3, 2008C): m/z: 904 ([M] ); elemental analysis
3
7.6 Hz, 2H; f), 6.88 (d, J(H,H) ˆ 8.2 Hz, 2H; d), 7.01 (s, 4H; a), 7.04 (dd,
calcd (%) for C46H42N4O12P2 (904.81): C 61.06, H 4.68, N 6.19; found: C
61.02, H 4.89, N 6.11.
3J(H,H) ˆ 8.2 Hz, J(H,H) ˆ 7.6 Hz, 2H; e), 7.40 (d, J(H,H) ˆ 8.2 Hz, 2H;
j), 7.43 (d, 3J(H,H) ˆ 8.2 Hz, 2H; i), 7.79 (s, 2H; h), 8.07 (t, 3J(H,H) ˆ
7.6 Hz, 1H; n), 8.20 (d, 3J(H,H) ˆ 7.6 Hz, 2H; m); 13C NMR (126 MHz,
CD3OD, 258C): d ˆ 36.4 (d, 1J(C, P) ˆ 134.4 Hz; 3), 41.8 (s; 10/17), 120.4 (s;
7), 122.7 (s; 5), 124.8 (s; 9), 126.3 (s; 12), 126.4 (s; 20), 130.6 (s; 8), 131.0 (s;
1), 131.1 (s; 15), 132.7 (s; 2), 134.3 (s; 14), 135.7 (s; 16), 140.9 (s; 21), 142.6
3
3
1,10,27,34-Tetraoxo-1,10-dioxido-1,10-diphospha-11,50-dioxa-26,35-di-
aza[2](3)benzeno[3](2)benzeno[1](3,4)nitrobenzeno[3](2)benzeno[3](3,5)-
nitrobenzeno[1](2)benzenocyclophane dilithium salt (2a): Cyclophane 8a
(187.7 mg, 0.207 mmol) was suspended in dry acetonitrile (2 mL) and
treated with a solution of lithium bromide (36.2 mg, 0.417 mmol) in dry
acetonitrile (2 mL). The mixture was heated to reflux for one week. Then
the product was filtered off and washed with cold acetonitrile and diethyl
ether. If the product still contained starting material, the whole procedure
was repeated with the appropriate additional amount of LiBr. Yield:
121.3 mg (0.137 mmol, 66%); m.p. 2508C (decomp, DSC); 1H NMR
(500 MHz, CD3OD): d ˆ 3.00 (d, 2J(H,P) ˆ 19.5 Hz, 4H; b), 4.01 (s, 4H; g),
6.83 (s, 2H; c), 6.95 (d, 3J(H,H) ˆ 7.6 Hz, 2H; f), 7.00 (d, 3J(H,H) ˆ 8.2 Hz,
2H; d), 7.13 (s, 4H; a), 7.19 (dd, 3J(H,H) ˆ 8.2 Hz, 3J(H,H) ˆ 7.6 Hz, 2H; e),
7.51 (dd, 3J(H,H) ˆ 8.2 Hz, 4J(H,H) ˆ 1.3 Hz, 2H; j), 7.55 (d, 3J(H,H) ˆ
(s; 11), 144.2 (s; 6), 149.9 (s; 13), 150.3 (s; 19), 154.9 (s; 4), 166.4 (s; 18); 31
P
À
NMR (202 MHz, CD3OD, 258C): d ˆ 18.8 (s); FT-IR: nÄ ˆ 3405 (N H),
ˆ
À
1672 (C O), 1605, 1586 (arene), 1527, 1348 (NO2), 1253 (P OAr), 1212
(P O) cmÀ1; MS (FAB ‡ NBA): m/z: 890 ([M‡H] ); elemental analysis
‡
À
calcd (%) for C43H35N5O12P2Li2 (889.61): C 58.06, H 3.97, N 7.87; found: C
55.36, H 4.23, N 7.33.
NMR host guest titrations: The guest compound was dissolved in the
appropriate amount of solvent and the resulting solution was evenly
distributed among 10 NMR tubes. The first NMR tube was sealed without
any guest. The host compound was also dissolved in the appropriate
amount of solvent and added in increasing amounts to the NMR tubes, so
that finally solutions with the following relative amounts (equiv) of host
versus guest compound were obtained: 0, 0.25, 0.50, 0.75, 1.00, 1.25, 1.50,
2.00, 3.00 and 5.00. The theoretical values (corrected for pipeting errors)
were afterwards again corrected by comparison with the integration of
various host and guest signals. All probes were measured with 256 pulses
per probe; all Dd values refer to the standard of the pure guest compound.
Ka and Ddsat were calculated by nonlinear regression from the observed Dd
values and the respective host and guest concentrations.
3
4
8.2 Hz, 2H; i), 7.64 (t, J(H,H) ˆ 7.9 Hz, 1H; o), 7.95 (d, J(H,H) ˆ 1.3 Hz,
2H; h), 8.06 (dd, 3J(H,H) ˆ 7.6 Hz, 4J(H,H) ˆ 1.3 Hz, 2H; n), 8.33 (t,
4J(H,H) ˆ 1.6 Hz, 1 H; m); 13C NMR (126 MHz, CD3OD, 258C): d ˆ 36.5
(d, 1J(C,P) ˆ 7.6 Hz; 3), 41.8 (s; 10), 42.1 (s; 17), 120.4 (s; 7), 122.7 (s; 5),
124.9 (s; 9), 126.2 (s; 12), 127.2 (s; 20), 130.4 (s; 22), 130.6 (s; 8), 131.0 (s; 1),
131.1 (s; 15), 132.2 (s; 21), 133.0 (s; 2), 134.2 (s; 14), 135.5 (s; 16), 136.1 (s;
19), 142.7 (s; 11), 144.1 (s; 6), 149.9 (s; 13), 154.9 (s; 4), 169.9 (s; 18); 31P
À
NMR (202 MHz, CD3OD): d ˆ 18.9 (s); FT-IR nÄ ˆ 3417 (N H), 1649
ˆ
À
(C O), 1605, 1585 (arom), 1528, 1347 (NO2), 1251 (P OAr), 1209
(P O) cmÀ1; MS (FAB ‡ NBA): m/z: 889 ([M‡H] ); elemental analysis
‡
À
A representative example is given below: 2a versus (R/S)-noradrenaline ¥
HCl (10) in CD3OD. Weighed amounts: 10: 1.78 mg in 8.00 mL; 2a:
11.74 mg in 0.61 mL; proton a: Kass [MÀ1] ˆ 1010 Æ 12%; Ddsat [ppm] ˆ
0.220 Æ 4%; proton b: Kass [MÀ1] ˆ 978 Æ 9%; Ddsat [ppm] ˆ 0.104 Æ 3%;
proton c: Kass [MÀ1] ˆ 1040 Æ 13%; Ddsat [ppm] ˆ 0.108 Æ 4%.
Self-association experiments: Solutions of 2a or 2b in D2O were prepared
in the following concentrations: 5 Â 10À3 m, 4 Â 10À3 m, 2 Â 10À3 m, 1 Â 10À3 m,
6 Â 10À4 m, 2 Â 10À4 m, 4 Â 10À5 m, and 8 Â 10À6 m. 1H NMR spectra were
recorded, starting with 256 pulses for the highest and finishing with 4096
pulses for the lowest concentration. The dilution curves obtained were
evaluated with standard nonlinear regression methods, and self-association
constants were calculated.
calcd (%) for C44H36N4O12P2Li2 (888.62): C 59.47, H 4.08, N 6.30; found: C
57.74, H 4.16, N 5.99.
1,10,27,34-Tetraoxo-1,10-dimethoxy-1,10-diphospha-11,50-dioxa-26,35-di-
aza[2](3)benzeno[3](2)benzeno[1](3,4)nitrobenzeno[3](2,1)pyridino[3](3,5)-
nitrobenzeno[1](2)benzenocyclophane (8b): A solution of 7b (0.620 g,
0.617 mmol) and triethylamine (0.69 mL, 4.94 mmol) in dry THF (50 mL)
and
a solution of 2,6-pyridinedicarboxylic acid dichloride (0.126 g,
0.617 mmol) in dry THF (50 mL) were continuously injected over 3 h at
room temperature into a mixture of dry THF (50 mL) and dry benzene
(10 mL). The reaction mixture was stirred overnight, then the precipitated
triethylammonium chloride was filtered off, the filtrate was evaporated to
dryness and the residue was chromatographed over silica gel 60 eluted with
chloroform/acetone ˆ 1:1 (Rf ˆ 0.33), producing a light yellow solid. Yield:
0.180 g (0.199 mmol, 32%); m.p. 2208C (decomp, DSC); 1H NMR
(500 MHz, CDCl3, 258C): d ˆ 3.27/3.28 (d (diastereomers), 2J(H,P) ˆ
20.2 Hz, 4H; b), 3.73/3.75 (d (diastereomers), 3J(H,P) ˆ 11.4 Hz, 6H; c),
Evaluation of the complex stoichiometry (Job plots): Complex stoichio-
metries were determined according to Job×s method of continuous
variations. Equimolar amounts of host and guest compound were dissolved
in the appropriate NMR solvent. These solutions were distributed among
11 NMR tubes in such a way that the molar fractions X of host and guest in
the resulting solutions increased (or decreased) from 0.0 to 1.0 (and vice
versa). The complexation-induced shifts (CIS) were multiplied by X and
plotted against X itself (Job plot).
3
3.95 (s, 4H; h), 4.85 (d, J(H,H) ˆ 6.3 Hz, 4H; l), 6.79 (s, 2H; d), 6.85 (d,
3J(H,H) ˆ 8.2 Hz, 2H; g), 6.90 (d, 3J(H,H) ˆ 7.6 Hz, 2H; e), 7.13/7.14 (dd
(diastereomers), 3J(H,H) ˆ 8.2 Hz, 3J(H,H) ˆ 7.6 Hz, 2H; f), 7.21 (s (dia-
3
4
stereomers), 4H; a), 7.42 (dd, J(H,H) ˆ 7.9 Hz, J(H,H) ˆ 1.3 Hz, 2H; k),
7.59 (d, 3J(H,H) ˆ 7.6 Hz, 2H; j), 7.80/7.82 (d (diastereomers), 4J(H,H) ˆ
Mass spectrometric measurements: ESI mass spectra were recorded on a
Finnigan MAT 95. Samples (20 mL) were introduced as 10À7 m solutions in
HPLC-grade methanol at flow rates of 20 mLminÀ1. Heated capillary
temperature: 1508C. Ion spray potential: 3.5 kV (positive ESI), 3.0 kV
(negative ESI). About 20 30 scans were averaged to improve the signal-
to-noise ratio.
3
3
1.3 Hz, 2H; i), 8.01 (t, J(H,H) ˆ 7.6 Hz, 1H; o), 8.30 (d, J(H,H) ˆ 7.6 Hz,
2H; n), 8.61 (t, 3J(H,H) ˆ 6.3 Hz, 2H; m); 13C NMR (126 MHz, CDCl3,
258C): d ˆ 32.8 (d, 1J(C,P) ˆ 139.9 Hz (diastereomers; 3), 40.7 (s; 11), 41.0
2
(s; 18), 53.2 (d, J(C,P) ˆ 3.0 Hz (diastereomers; 4), 118.6 (s; 10), 120.8 (s;
6), 125.0 (s; 13), 125.5 (s; 8), 129.7 (d, 2J(C,P) ˆ 2.4 Hz (diastereomers; 2),
130.0 (s; 16), 130.2 (s; 1), 131.0 (s; 15), 132.3 (s; 9), 134.6 (s; 17), 139.1 (s;
21), 141.1 (s; 12), 142.1 (s; 22), 142.2 (s; 7), 148.6 (s; 14), 148.6 (s; 20), 150.8
Molecular modeling: Force-field calculations were initially carried out as
molecular mechanics calculations without solvent (Cerius2, Molecular
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1497 $ 17.50+.50/0
1497

T. Schrader et al.
FULL PAPER
Simulations 1997, force field: Dreiding 2.21). To establish the minimum
energy conformation of the free host and guest molecule as well as their 1:1
complex, Monte Carlo simulations were carried out in water (Macro-
Model 7.0, Schrˆdinger Inc., 2000. Force-field: Amber*). A 3000-step
Monte Carlo simulation was carried out, followed by a molecular dynamics
calculation for 10 ps at 300 K. The three best structures for the 1:1 complex
varied in their total energy by only 5 kJmolÀ1. In each case, the adrenaline
molecule is included in the cavity of the macrocycle with specific
recognition of the amino alcohol by the bisphosphonate anions. A beautiful
sandwich arrangement is consistently formed between the catechol and the
double nitrobenzene wall in 2a or b. In addition, the upper of both phenolic
OH groups forms two hydrogen bonds to the isophthaloyl amides. The
whole complex structure is slightly twisted in order to minimize bond and
torsional strains and maximize van der Waals interactions. One of the
anionic phosphonate O atoms is rotated outwards to gain solvation energy
with the solvent. In chloroform or without solvent both anionic O atoms are
predicted to form strong hydrogen bonds with the ammonium cation of the
guest.
[10] X. Huang, B. H. Rickman, B. Borhan, N. Berova, K. Nakanishi, J. Am.
Chem. Soc. 1998, 120, 6185 6186.
[11] a) T. Schrader, Angew. Chem. 1996, 108, 2816 2818; T. Schrader,
Angew. Chem. Int. Ed. Engl. 1996, 35, 2649 2651; b) T. Schrader, J.
Org. Chem. 1998, 63, 264 272.
[12] M. I. Rodriguez-Franco, P. San Lorenzo, A. Martinez, P. Navarro,
Tetrahedron 1999, 55, 2763.
[13] K. Odashima, K. Yagi, K. Tohda, Y. Umezawa, Bioorg. Med. Chem.
Lett. 1999, 9, 2375.
[14] R. R. Makote, M. M. Collinson, Chem. Mater. 1998, 10, 2440.
[15] A. R. Bernardo, J. F. Stoddart, A. Kaifer, J. Am. Chem. Soc. 1992, 114,
10624 10631.
[16] J. A. Gavin, M. E. Garcia, A. J. Benesi, T. E. Mallouk, J. Org. Chem.
1998, 63, 7663 7669.
[17] a) R. Kuhn, F. Erni, T. Bereuter, J. H‰usler, Anal. Chem. 1992, 64,
2815 2820; b) R. Kuhn, C. Steinmetz, T. Bereuter, P. Haas, F. Erni, J.
Chromatogr. A 1994, 666, 367 373; c) W. X. Huang, H. Xu, S. D.
Fazio, R. V. Vivilecchia, J. Chromatogr. B 1997, 695, 157 162.
[18] a) S. M. Strawbridge, S. J. Green, J. H. R. Tucker, Chem. Commun.
2000, 2393 2394; b) S. M. Strawbridge, S. J. Green, J. H. R. Tucker,
Phys. Chem. Chem. Phys. 2000, 2, 2393 2394.
[19] a) S. F. Hou, H. Q. Fang, H. Y. Chen, Anal. Lett. 1998, 31, 765 775;
b) S. F. Hou, K. S. Yang, H. Q. Fang, H. Y. Chen, Chin. Chem. Lett.
1998, 9, 165 168.
[20] M. Morita, O. Niwa, T. Horiuchi, Electrochim. Acta 1997, 42, 3177
3183.
[21] F. P. Schmidtchen, Z. Naturforsch. C. Biosci. 1987, 42, 476 485.
[22] E. Kimura, H. Fujioka, M. Kodama, J. Chem. Soc. Chem. Commun.
1986, 1158 1159.
[23] K. Hayakawa, K. Kido, K. Kanematsu, J. Chem. Soc. Perkin Trans. 1
1988, 511 519.
[24] C. E. Park, Y. G. Jung, J. I. Hong, Tetrahedron Lett. 1998, 39, 2353
2356.
[25] a) M.-F. Paugam, L. S. Valencia, B. Bogess, B. D. Smith, J. Am. Chem.
Soc. 1994, 116, 11203; b) M.-F. Paugam, J. T. Biens, B. D. Smith, A. J.
Christoffels, F. de Jong, D. N. Reinhoudt, J. Am. Chem. Soc. 1996, 118,
9820 9825.
[26] M. B. Inoue, E. F. Velazquez, M. Inoue, Q. Fernando, J. Chem. Soc.
Perkin Trans. 2 1997, 2113 2118.
FT-IR experiments: FT-IR spectra were recorded for noradrenaline
hydrochloride, free host 2a, and the corresponding 1:1 complex (KBr).
The IR spectrum of the complex was identical to the addition of host and
guest spectra, with two exceptions:
À
a) the strong band for the host×s anionic P O bond (symmetrical and
asymmetrical vibrational stretch) shifted from 1070.8 to 1060.2 cmÀ1
and from 1210.8 to 1186.3 cmÀ1 (D ˆ 8.6 and 24.5 cmÀ1). This shift to
lower values of nÄ corresponds to hydrogen bond formation with
noradrenaline×s ammonium functionality;
ˆ
b) the host×s strong amide carbonyl band (C O, vibrational stretch) shifted
from 1650.7 to 1646.5 cmÀ1 (D ˆ 4.2 cmÀ1). This corresponds to the
À
formation of hydrogen bonds between its coupled N H bond and the
phenolic hydroxy group.
1
NOESY experiments: Standard H NOESY experiments were performed
with the host 2a and 2b as well as their complexes with noradrenaline ¥ HCl
on the Bruker 500 MHz NMR spectrometer. Mixing times varied between
100 and 200 ms. In spite of its high viscosity leading to enforced spin
lattice relaxation we chose [D6]DMSO as solvent, because here the
important amide proton NOEs could be measured. Positive reciprocal
NOE peaks were classified as w ˆ weak, m ˆ medium, and s ˆ strong.
[27] C. Mannironi, C. Scerch, P. Fruscoloni, G. P. Tocchini-Valentini, RNA
2000, 6, 520 527.
[28] M. Herm, T. Schrader, Chem. Eur. J. 2000, 6, 47 53.
[29] M. Herm, O. Molt, T. Schrader, Angew. Chem. 2001, 113, 3148 3151;
Angew. Chem. Int. Ed. 2001, 40, 3244 3248.
Acknowledgements
[30] S. Trumpp-Kallmeyer, J. Hoflack, A. Bruinvels, M. Hibert, J. Med.
Chem. 1992, 35, 3448 3462.
This work was supported by the Deutsche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie.
[31] a) M. Mammen, E. I. Shakhnovich, J. M. Deutch, G. M. Whitesides, J.
Org. Chem. 1998, 63, 3821 3830; b) D. A. Jeyaraj, V. K. Yadav, M.
Parvez, H. M. Gauniyal, J. Org. Chem. 1998, 63, 3831 3837.
[32] For selected leading references, see: a) M. Dhaenens, L. Lacombe, J.-
M. Lehn, J.-P. Vigneron, J. Chem. Soc. Chem. Commun. 1984, 1097
1099; b) J. Jazwinski, J.-M. Lehn, D. Lilienbaum, R. Ziessel, J.
Guilhem, C. Pascard, J. Chem. Soc. Chem. Commun. 1987, 1691;
c) H.-J. Schneider, T. Blatter, S. Simova, I. Theis, J. Chem. Soc. Chem.
Commun. 1989, 580 581; d) H.-J. Schneider, D. Ruf, Angew. Chem.
1990, 102, 1192 1194; Angew. Chem. Int. Ed. Engl. 1990, 29, 1159
1160; e) P. Scrimin, P. Tecilla, U. Tonnelato, M. Vignana, J. Chem. Soc.
Chem. Commun. 1991, 449; f) T. H. Webb, H. Suh, C. S. Wilcox, J. Am.
Chem. Soc. 1991, 113, 8554 8555; g) H.-J. Schneider, Chem. Soc. Rev.
1994, 227 234; h) J.-M. Lehn, R. Meric, J.-P. Vigneron, M. Cesario, J.
Guilhem, C. Pascard, Z. Asfari, J. Vicens, J. Supramol. Chem. 1995, 5,
97 103.
[33] M. Kamieth, F.-G. Kl‰rner, F. Diederich, Angew. Chem. Int. Ed. 1998,
37, 3303 3306.
[34] G. Zigeuner, E. Fuchs, H. Brunetti, H. Sterk, Monatsh. Chem. 1966, 97,
36.
[35] V. Mancini, G. Marino, L. Giachetti, Gazz. Ital. 1974, 104, 549.
[36] Organolithium compounds: H. Gilman, L. Santucci, D. R. Swayam-
pati, R. O. Ranck, J. Am. Chem. Soc. 1957, 79, 3077.
[37] Organolithium cuprates: E. J. Corey, G. H. Posner, J. Am. Chem. Soc.
1967, 89, 3911; for diarylmethanes: B. J. Whitlock, H. W. Whitlock, J.
Org. Chem. 1980, 45, 12.
[1] T. P. Iismaa, T. J. Biden, J. Shine, G Protein-Coupled Receptors,
Springer, Heidelberg, 1995.
[2] C. R. Craig, R. E. Stitzel, Modern Pharmacology, Little, Brown,
Boston, 1990.
[3] S. B. Ligett, Pharmacology 2000, 61, 167 173.
[4] M. G. Bock, M. A. Patane, Annu. Rep. Med. Chem. 2000, 35, 221 230.
[5] a) M. R. Bristow, Circulation 2000, 101, 558 569; b) I. B. Squire, D. B.
Barnett, Br. J. Clin. Pharmacol. 2000, 49, 1 9; c) A. Himmelmann,
Cardiovasc. Drugs Ther. 1999, 13, 469 477.
[6] Recent GPCR crystal structures: a) E. Pebay-Peyroula, G. Rummel,
J.-P. Rosenbusch, E. M. Landau, Science 1997, 277, 1676; b) D. Doyle,
J. M. Cabral, R. A. Pfuetzner, A. Kuo, J. M. Gulbis, S. L. Cohen, B. T.
Chait, R. Mackinnon, Science 1998, 280, 69; c) K. Palczewski, T.
Kumasaka, T. Hori, C. A. Behnke, H. Motoshima, B. A. Fox, I.
Le Trong, D. C. Teller, T. Okada, R. O. Stenkamp, M. Yamamoto, M.
Miyano, Science 2000, 289, 739 745.
[7] a) J.-P. Behr, J.-M. Lehn, P. Vierling, Helv. Chim. Acta 1982, 65, 1853
1867; b) E. Kimura, A. Watanabe, M. Kodama, J. Am. Chem. Soc.
1983, 105, 2063.
[8] L. Campayo, J. M. Bueno, C. Ochoa, P. Navarro, J. Jimenez-Barbero,
G. Pepe, A. Samat, J. Org. Chem. 1997, 62, 2684.
[9] T. Ishizu, J. Hirayama, S. Noguchi, Chem. Pharm. Bull. 1994, 42,
1146 1148.
1498
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1498 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 6

Receptor for Adrenaline
1485 1499
[38] M. Gaudemar, Bull. Soc. Chim. Fr. 1962, 974 987.
[39] S. C. Berk,M. C. P. Yeh, N. Jeong, P. Knochel, Organometallics 1990, 9,
3053 3064.
[40] a) E. Negishi, A. O. King, N. Okukado, J. Org. Chem. 1977, 42, 1821
1823; b) Reviews: Organozinc reagents (Eds.: P. Knochel, P. Jones),
Oxford University Press, Oxford, 1999, pp. 213 243; Metal-Catalyzed
Cross-Coupling Reactions (Eds.: F. Diederich, P. J. Stang), Wiley-
VCH, Weinheim, 1998, pp. 1 47.
[45] R. P. Bonar-Law, A. P. Davis, B. A. Murray, Angew. Chem. 1990, 102,
1497 1499; Angew. Chem. Int. Ed. Engl. 1990, 29, 1407 1408; S.
Anderson, U. Neidlein, V. Gramlich, F. Diederich, Angew. Chem.
1995, 107, 1722 1724; Angew. Chem. Int. Ed. Engl. 1995, 34, 1596
1600.
[46] U. Neidlein, F. Diederich, Chem. Commun. 1996, 1493.
[47] a) P. Job, C. R. 1925, 180, 928; b) M. T. Blanda, J. H. Horner, M.
Newcomb, J. Org. Chem. 1989, 54, 4626.
[41] P. Knochel, personal communication. We thank Prof. Knochel for this
valuable advice.
[42] a) K. Saigo, R.-J. Lin, M. Kubo, A. Youda, M. Hasegawa, J. Am.
Chem. Soc. 1986, 108, 1996; b) K. Saigo, N. Kihara, Y. Hashimoto, R.-
J. Lin, H. Fujimura, Y. Suzuki, M. Hasegawa, J. Am. Chem. Soc. 1990,
112, 1144.
[43] a) R. Karaman, A. Goldblum, E. Breuer, J. Chem. Soc. Perkin Trans.
1989, 765; b) H. Krawczyk, Synth. Commun. 1997, 27, 3151 3161.
[44] A. P. Davis, R. Wareham, Angew. Chem. 1998, 110, 2397 2401;
Angew. Chem. Int. Ed. 1998, 37, 2270 2273.
[48] Since both the NOEs between OH and NH protons as well as the
diagonal are negative, we cannot a priori exclude exchange phenom-
ena. However, no cross-peak was found between water and the NH
proton.
[49] a) H. J. Schneider, R. Kramer, S. Simova, U. Schneider, J. Am. Chem.
Soc. 1988, 110, 6442; b) C. S. Wilcox in Frontiers in Supramolecular
Chemistry (Ed.: H. J. Schneider), VCH, Weinheim, 1991, p. 123.
[50] J. C. Ma, D. A. Dougherty, Chem. Rev. 1997, 97, 1303 1324; J. P.
Gauivan, D. A. Dougherty, J. Am. Chem. Soc. 2000, 122, 870 874.
Received: September 3, 2001 [F3530]
Chem. Eur. J. 2002, 8, No. 6
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0806-1499 $ 17.50+.50/0
1499