A selective biomimetic tweezer for noradrenaline

Article abstract of DOI:10.1021/ja035212l

The new highly preorganized tweezer molecule 1 binds noradrenaline in polar solvents with unprecedented specificity. It uses a biomimetic recognition pattern and rejects almost all other neurotransmitters. LB experiments on a film balance reflect the same

Full text of DOI:10.1021/ja035212l

Published on Web 09/12/2003
A Selective Biomimetic Tweezer for Noradrenaline
O. Molt, D. Ru¨beling, and T. Schrader*
Philipps-UniVersita¨t Marburg, Fachbereich Chemie, Hans-Meerwein-Strasse, 35032 Marburg
Received March 18, 2003; E-mail: schradet@mailer.uni-marburg.de
Catecholamines such as (nor-)adrenaline or dopamine are
important hormones and neurotransmitters. Through agonistic
interaction with adrenergic receptors a broad range of vital body
functions are influenced, and these receptors are still hot targets
for pharmaceutical research.1 Many groups have been inspired to
design artificial host molecules for catecholamines. However, most
of these systems are monotopic and nonspecific.2
We intended to shape a highly preorganized receptor molecule,
similar to those in Nature, for a more favorable complexation
entropy and improved desolvation of the included guest. Tweezer
molecules with stiff elements and connections have served this
purpose very successfully in recent years.3
Figure 1 demonstrates the biomimetic design of this new host:
Similar to the natural example, the ammonium alcohol is bound
by salt bridges and π-cation interactions, reinforced by a network
of ionic hydrogen bonds. The catechol itself is buried in an aromatic
cleft capable of π-stacking, and the phenolic hydroxyls are all extra
hydrogen-bonded.4
Figure 1. Schematic binding pattern of noradrenaline in the natural
â-adrenergic receptor (left) and in the rigid biomimetic host 1 (right).
In contrast to macrocyclic hosts which require a multistep
synthesis with the critical macrocyclization in the end, the new
receptor is accessible in analytically pure form by a short and
convergent synthesis: biphenylic side walls are linked to an
aromatic diaminophenazine headgroup by standard amide coupling,
followed by deprotection of the other end and attachment of both
phosphonate moieties.5 Apart from the rotations of the biphenyls
around their own axis, this scaffold should exist in a quite immobile
conformation. Furthermore, the well-known formation of intramo-
lecular hydrogen bonds between both amide protons and the
phenazine ring-nitrogen fixes the scaffold at an aromatic distance
of 7.0 Å and orientates the amide protons inward, suitable for
catechol recognition.6 At the opposite part of the biphenyls two
phosphonate anions guarantee the strong and essential recognition
of the ammonium alcohol moiety.
Figure 2. Lewis-structures of tweezers 1-3 and association constants of
their complexes with noradrenaline hydrochloride in d4-methanol at 20 °C.
To validate the chosen building blocks we synthesized three
different tweezers and examined their complexes with noradrenaline
hydrochloride. Job plots revealed that in all cases 1:1 complexes
are formed, and the association constants in methanol determined
1
by H NMR titrations confirmed our strategy (Figure 2).7 While
host 2 with much too long tolane side walls does not bind
noradrenaline at all, the biphenylic receptor 3 forms a weak complex
(Ka ) 120 M-1). Replacing the flexible m-xylylene headgroup with
the rigid phenazine moiety in receptor 1 strongly improves the
affinity for the desired guest (Ka ) 1800 M-1).
In complexes of adrenaline derivatives with 1, all catecholamine
protons undergo complexation-induced shifts, demonstrating the
close proximity of host and guest.8
Figure 3. (Left) ESI mass spectrum of a solution of receptor 1 and
noradrenaline hydrochloride (NA) in methanol. Host peaks: m/z ) 759 [1
+ 3H]+, 781 [1 + 2H + Na]+, 803 [1 + H + 2Na]+, 825 [1 + 3Na]+.
Complex peaks: m/z ) 928 [NA + 1 + 2H]+, 940 [NA + 1 + 2Li]+.
Right: Complex between receptor 1 and noradrenaline according to Monte
Carlo simulations in water with subsequent molecular dynamics (right).
To our disappointment, no shifts were found in the UV/vis spectra
of 1 on addition of noradrenaline.
The effective 1:1 complex formation between 1 and noradrenaline
could also be monitored by ESI-MS, producing clean mass spectra
with host and aggregate ion peaks, exclusively (Figure 3).9 In the
FT-IR, a strong shift (40 cm-1) of the asymmetric PdO valence to
lower wavenumbers demonstrates the amino alcohol recognition.
Molecular modeling studies strongly support our design (Figure
3).10 Furthermore, molecular dynamics calculations confirm the
rigidity of the upper part of the host and the flexibility of the
phosphonates. Even inside the cavity the guest has some mobility,
indicating that it is large enough to include the guest.
9
12086
J. AM. CHEM. SOC. 2003, 125, 12086-12087
10.1021/ja035212l CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Association Constants [M-1] from NMR Titrations of 1
found in Table 1 into the aqueous subphase (10-4 M) leads to strong
effects reflecting the interaction with the embedded receptor
molecule (no effects are produced with stearic acid alone). By far
the largest shift is obtained from noradrenaline, followed by much
smaller shifts from adrenaline and dopamine. All other guests
produce negligible shifts, which beautifully supports the high
selectivity of 1 for noradrenaline (Figure 4).
We are currently developing highly selective sensor devices in
joint projects. To this end, host 1 will be incorporated in lipid/
polydiacetylene assemblies for a photometric detection of norad-
renaline.13 Alternatively, a titanium dioxide matrix including 1 as
receptor site will be immobilized on electrodes (ITO) or field-effect
transistors for electrochemical catecholamine detection.14
with Neurotransmitters and Related Guests in d4-Methanol at 25
°Ca
K
K
a
a
entry
guest hydrochloride
[M-1
]
entry
guest hydrochloride
[M-1
]
1
2
3
4
5
6
(R/S)-noradrenaline 1800
7
8
9
(L)-tryptophan methyl ester 240
(R/S)-adrenaline
dopamine
phenethylamine
2-aminoethanol
catechol
260
340
(L)-tyrosine methyl ester
(R)-propranolol
170
130
<1
<1
<1b 10 GABA
<1
<1
11 glycine
a Errors in Ka are standard deviations from the nonlinear regressions and
were estimated at (10-45%. b Lowest detection limits.
Acknowledgment. This work was supported by the Deutsche
Forschungsgemeinschaft.
Supporting Information Available: Synthetic details, NMR bind-
ing experiments, Langmuir film experiments, and molecular modeling
data (PDF). This material is available free of charge via the Internet at
http://pubs.acs.org.1
References
(1) (a) Ligett, S. B. Pharmacology 2000, 61, 167-173. (b) Bock, M. G.;
Patane, M. A. Annu. Rep. Med. Chem. 2000, 35, 221-230.
(2) For an overview, see: Schrader, T.; Herm, M.; Molt, O. Chem. Eur. J.
2002, 8, 1485-1499.
Figure 4. (Left) Schematic representation of the tweezer 1 embedded in a
stearic acid monolayer at the air/water interface. The cavity is open to the
water sub-phase and received a catecholamine guest molecule. (Right)
Pressure-area isotherms of stearic acid (S) and receptor in a monolayer
over water (S + R), dopamine (1), adrenaline (2), or noradrenaline (3).
(3) (a) Zimmerman, S. C. Top. Curr. Chem. 1993, 165, 71-102. (b) Brown,
S. P.; Schaller, T.; Seelbach, U. P.; Koziol, F.; Ochsenfeld, C.; Kla¨rner,
F.-G.; Spiess, H. W. Angew. Chem., Int. Ed. 2001, 40, 717-720. (c) Krebs,
F. C.; Jorgensen, M. J. Org. Chem. 2001, 66, 6169-6173.
(4) For complexation through charged hydrogen bonds, see: Echavarren, A.;
Galan, A.; Lehn, J.-M.; de Mendoza, J. J. Am. Chem. Soc. 1989, 111,
4994-4995; Snowden, T. S.; Anslyn, E. V. Curr. Opin. Chem. Biol. 1999,
3, 740-746; Berger, M.; Schmidtchen, F. P. J. Am. Chem. Soc. 1999,
121, 9986-9993. For catechol binding with cleft-type hosts, see:
Sijbesma, R. P.; Nolte, R. J. Top. Curr. Chem. 1995, 175, 25-26.
(5) All synthetic details are found in the Supporting Information.
(6) Safarowsky, O.; Nieger, M.; Fro¨hlich, R.; Vo¨gtle, F. Angew. Chem., Int.
Ed. 2000, 39, 1616-1618; Zimmerman, S. Bioorg. Med. Chem. 1996, 4,
1107.
A thorough examination of the tweezer’s binding selectivity was
carried out by titrating structurally related guests with 1 (Table 1).
Representative guests are hydrochlorides of amines, amino alcohols,
and amino acids with only marginal structural deviations from
noradrenaline.11
Systematic alterations of noradrenaline’s structure in minute steps
(single methyl and OH groups) lead to a noticeable decrease in Ka
at every step, demonstrating the specificity of receptor 1 for this
guest (entries 1-4). Cutting the guest in two halves completely
eliminates any affinity toward 1 (entries 5,6). Only replacement of
the catechol by larger electron-rich π faces as in aromatic amino
acid esters (entries 7,8) or the adrenergic receptor antagonist
propranolol (entry 9) restores a weak attraction by host 1. Molecular
modeling studies reveal a distorted complex geometry for these
cases, because their additional substituents prevent the inclusion
of amino acids and â-blockers inside the cavity of 1. None of the
ammonium-based neurotransmitters listed in entries 10-11 showed
any interaction with 1. These investigations characterize 1 as an
artificial receptor molecule with high noradrenaline specificity.
The highly amphiphilic structure of 1 prompted us to incorporate
the receptor molecule in a stearic acid monolayer at the air/water
interface. In the Langmuir film balance, substantial shifts are
produced in the pressure/area diagram, indicating rapid incorpora-
tion of 1 into the monolayer.12 Subinjection of various analytes
(7) Wilcox, C. S. In Frontiers in Supramolecular Chemistry; Schneider, H.
J., Ed.; Verlag Chemie: Weinheim, 1991; p 123.
(8) The tendency of 1 to self-associate in methanol was determined at Ksa e
50 M-1 by dilution titration.
(9) No molecular ion peaks were found for host dimers or higher aggregates.
(10) MacroModel 7.0, Schro¨dinger Inc., Force-Field: Amber*, Monte Carlo
simulations in water: 3000 steps.
(11) Due to the low methanol solubility of 1, the NMR titrations had to be
carried out at millimolar concentrations, and the amount of added host
did often not surpass 2 equiv. At these conditions, however, the
complexation-induced shifts remained rather small, and only in the case
of noradrenaline was 80% saturation reached. In all other cases, saturation
did not exceed 50%, owing to their weak affinity towards 1. To secure
the high Ka value found for noradrenaline, we repeated the respective
binding experiment and obtained similar association constants from two
independent proton signals.
(12) For a recent review, see: Ariga, K.; Kunitake, T. Acc. Chem. Res. 1998,
31, 371-378.
(13) Kolusheva, S.; Shahal, T.; Jelinek, R. J. Am. Chem. Soc. 2000, 122, 776-
780.
(14) Lahav, M.; Kharitonov, A. B.; Willner, I. Chem. Eur. J. 2001, 7,
3992-3997.
JA035212L
9
J. AM. CHEM. SOC. VOL. 125, NO. 40, 2003 12087