A combined experimental and theoretical study of the pH-dependent binding mode of NAD+ by water-soluble molecular clips

Article abstract of DOI:10.1002/poc.1519

The highly selective recognition process of NAD⁺ and NADH (as important cofactors of many redox enzymes) by molecular clips in aqueous solution is studied systematically by a combined experimental and quantum-chemical approach. The strongly pH-

Full text of DOI:10.1002/poc.1519

Research Article
Received: 1 October 2008,
Revised: 19 November 2008,
Accepted: 27 November 2008,
Published online in Wiley InterScience: 13 March 2009
(www.interscience.wiley.com) DOI 10.1002/poc.1519
A combined experimental and theoretical
study of the pH-dependent binding mode of
NADþ by water-soluble molecular clips
Jolanta Polkowskaa, Frank Bastkowskia, Thomas Schradera,
a
b
by
*
Frank-Gerrit Kla¨rner , Jan Zienau , Felix Koziol
and Christian Ochsenfeldb
The highly selective recognition process of NADR and NADH (as important cofactors of many redox enzymes) by
molecular clips in aqueous solution is studied systematically by a combined experimental and quantum-chemical
1
approach. The strongly pH-dependent complexation-induced H NMR shifts of the guest molecule indicate that in
buffered aqueous solution at pH ¼ 7.2 the nicotinamide ring, the active site of NADR, is preferentially bound inside the
cavity of the molecular clip, whereas in pure water under slightly acidic conditions both units (the nicotinamide ring as
well as the adenine moiety) are located outside the cavity. The latter finding is explained by a competing self-
aggregation of NADR which prohibits the recognition process. In addition, the investigation of the NADR fragments
NMNA, NMN, and AMP as well as the comparison of measured and computed chemical shieldings provides
information on possible binding modes. Under equal conditions the binding of NADH to the molecular clip is
significantly weaker than that of NADR, so that the oxidation states (NADR/NADH) can be distinguished by the
molecular clips. The nucleotides NMN and AMP are bound less strongly by the molecular clips than NADR. The weaker
binding indicates that multiple aromatic pSp and cationSp host–guest interactions only possible in NADR have a
synergetic effect on the complex stability. In addition to the inhibition of the cofactor NADR, a further implication is
the development of sensors since a quenching of fluorescence is observed for specific molecular clips by the addition
of NADR. Copyright ß 2009 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: molecular clips; enzyme cofactors; nucleotide; supramolecular chemistry
molecular clips 1b and 1d substituted with lithium phosphonate
INTRODUCTION
or phosphate groups with its active S-methylsulfonium side-chain
inside the clip cavity just as predicted.[6] However, in the
host–guest complexes of NAD(P)þ with the molecular clip 1a
(substituted by tetra-n-butylammonium methanephosphonate
groups in the central benzene spacer-unit) the comparison of the
experimental and calculated 1H NMR shifts of the guest protons
clearly indicates that neither the nicotinamide ring nor the
adenine unit of NAD(P)þ is included inside the clip cavity in
slightly acidic aqueous solution.[4,5] In this work, we focus on the
structures and stabilities of the host–guest complexes of clips
Many enzymes use cofactors to catalyze, inter alia, redox reactions,
decarboxylations, or methyl transfer reactions.[1,2] The biologically
active part of these cofactors comprises an electron-poor, often
cationic heterocyclic ring system or a trialkylsulfonium salt. Direct
external interference with the catalytic process hence requires
blocking of the cofactor site, e.g., by synthetic receptor molecules.
A perfect design for this purpose is a rigid molecular clip with an
electron-rich interior for cation inclusion and solubilizing
functional groups at its periphery.[3] As a prototype, host family
1a–d with naphthalene sidewalls and a central benzene spacer
unit provide an extremely electron-rich cavity and retain
water-solubility due to their pendant methanephosphonate,
hydrogenphosphate, or phosphate groups (Scheme 1). It is
therefore expected that these artificial cofactor traps chemose-
lectively bind their guests in aqueous solution by embracing their
most electron-deficient sites and thereby affect enzymatic
reactivity. Recently, we reported the synthesis of 1a–d and the
complex formation of 1a as host molecule with various
pyridinium, thiazolium, and sulfonium salts such as
N-methylnicotinamide iodide (NMNA 3), nicotinamide adenine
dinucleotide (NADþ 2), and thiamine diphosphate (TPP) as guest
molecules.[4–9] Even non-aromatic S-adenosylmethionine (SAM)
carrying a tetrahedral sulfur cation was found to be bound by the
*
Correspondence to: F.-G. Kla¨rner, Institut fu¨r Organische Chemie, Universita¨t
Duisburg-Essen, 45117 Essen, Germany.
E-mail: frank.klaerner@uni-duisburg-essen.de
a
b
J. Polkowska, F. Bastkowski, T. Schrader, F.-G. Kla¨rner
Institut fu¨r Organische Chemie, Universita¨t Duisburg-Essen, 45117 Essen,
Germany
J. Zienau, F. Koziol, C. Ochsenfeld
Institut fu¨r Physikalische Chemie und Theoretische Chemie, Universita¨t
Tu¨bingen, 72076 Tu¨bingen, Germany
y
Present address: Department of Chemistry and Chemical Biology, Harvard
University, Cambridge, MA 02138, USA.
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.

J. POLKOWSKA ET AL.
1b–d substituted by lithium methanephosphonate or phosphate
groups with NADþ 2, NMNA 3, and the nucleotides NMN 4, AMP 5,
and NADH 6 as guest molecules in buffered aqueous solution at
neutral pH value. Under these conditions either the nicotinamide
ring or the adenine unit of these guest molecules is, indeed,
bound inside the clip cavity. NMNA 3, NMN 4, and AMP 5
represent fragments of the NADþ-molecule whose behavior
toward the molecular clips yields additional information about the
binding of NADþ 2 within clips 1b–d. The experimental results are
supported by quantum-chemical calculations of complexatio-
n-induced NMR chemical shifts of the guest protons which allow
us to gain further insight into the binding modes of these
host–guest complexes. With these findings a new door is opened
for the development of a potential supramolecular control of
enzymatic reactions. First experiments with the clips 1b and 1d
and the enzymes alcohol dehydrogenase (AD) and glucose-6-
phosphate dehydrogenase (G6PD) using NAD(P)þ as cofactors
look very promising.[10,11]
RESULTS AND DISCUSSION
ITC and NMR measurements: self-assembly of NADR and
molecular clips
Before discussing the host–guest binding of NADþ 2, NMNA 3,
and the nucleotides 4–6 by the molecular clips 1b–d, it is
important to describe the individual properties of these
compounds in aqueous solution: the pH-dependent self-
aggregation of the nucleotides (NADþ and AMP) and the
formation of self-assembled clip dimers.
Commercially available NADþ 2 dissolved in pure water reacts
slightly acidic (pH ¼ 3.3) and is, presumably, protonated at
nitrogen atom N-1 of the adenine unit.[12]. Dilution experiments
followed by isothermal titration calorimetry (ITC) provide good
evidence that this form of NADþ aggregates in aqueous solution
(Figure 1(a)).[5] Attempts to fit the experimental heat evolved
during the dilution of NADþ with a curve calculated for the
Figure 1. Plot of the heat pulses measured during the dilution of NADþ 2: 30 ꢀ 10 ml portions of a solution of NADþ ([NADþ] ¼ 10 mM) (a) in H2O, (c) in
aqueous buffer were added to H2O or buffer (each 1.4211 ml). (b) and (d) concentration dependence of the heat development (in mcal) per injection. The
solid line in (b) represents the curve calculated with the assumption that a potential self-assembled dimer (NADþ)2 is formed. The missing fit of the
calculation with the experimental data indicates a stoichiometry of the self-aggregate higher than that of a dimer
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790

A STUDY OF THE PH-DEPENDENT BINDING MODE OF NADþ
assumed formation of a self-assembled dimer (NADþ)2 failed
(Figure 1(b)). This indicates that protonated NADþ forms a higher
self-aggregate in water.[13] In buffered aqueous solution at
pH ¼ 7.2 no heat is evolved during the dilution of NADþ
(Figure 1(c) and (d)) showing that deprotonated NADþ 2 exists as
monomer under these conditions. Similar results by the use of ITC
were obtained for AMP 5. In pure water it forms an aggregate
similar to NADþ whereas in buffered solution it exists as
monomer. In the case of NMN 4, no experimental evidence has
been found by ITC for self-association under both conditions
suggesting that NMN exists in its monomeric form in pure water,
as well. (The experimental data of the ITC measurements of AMP
and NMN are shown in the Supporting Information).
buffered and in pure aqueous solution by 1H NMR spectroscopy.
The maximum complexation-induced 1H NMR shifts of the guest
protons, Ddmax (Ddmax ¼ d0 ꢁ dHG), and the association constants,
1
Ka, were determined by the methods of H NMR titration, from
the dependence of the observed complexation-induced 1H NMR
shifts, Ddobs (Ddobs ¼ d0 ꢁ dobs), of the guest protons either on
dilution of the host concentration at constant guest concen-
tration or on dilution of a solution containing both host and guest
in ca. 1:1 ratio.[15]
Host–guest complexes with NADR
2
1
Large Ddmax values of the guest protons were found in the H
In pure water, the phosphonate-substituted clip 1b reacts
almost neutral (pH ¼ 7.2), whereas the hydrogenphosphate-
substituted clip 1c reacts slightly acidic (pH ¼ 4.5, [1c] ¼ 3.77 mM)
and the phosphate clip 1d slightly basic (pH ¼ 8.5,
[1d] ¼ 5.30 mM). In buffered solution at pH ¼ 7.2 or 7.6 1c is
partially deprotonated and 1d partially protonated leading to
one and the same compound independently from the starting
material. The notation ‘clip 1d’ in Tables 1–4 means that 1d was
NMR titration experiments for the complex formation of NADþ
2
with the clips 1b and 1d in buffered aqueous solution at pH ¼ 7.2
(Table 1). In pure water under slightly acidic condition, however,
only small Ddmax values of the corresponding NADþ protons were
observed in the complexes of clips 1b and 1c (Table 1)
analogously to those found for the NADþ complex of clip 1a.[5]
These findings allow the following conclusions: in buffered
solution the nicotinamide ring of NADþ 2 is bound inside the
cavity of clip 1b or 1d comparable to the corresponding
complexes of NMNA 3 (vide infra). The relatively large Ddmax
values also observed for the adenine protons of 2 (Table 1)
indicate that the adenine unit may be also positioned inside the
clip cavity. A possible explanation for the observed shifts could be
the existence of a shuttle process in a (1:1) complex occurring
from the nicotinamide ring to the adenine unit, presumably via
mutual dissociation/association, fast with respect to the ‘‘NMR
timescale,’’ so that all NMR signals are averaged.[16,17] Further
insight into these intriguing complex structures comes from
quantum-chemical 1H NMR shift calculations (vide infra). The
small Ddmax values observed for all NADþ protons in the
complexes of 1b and 1c in pure water clearly indicate that in
these cases neither the nicotinamide ring nor the adenine unit is
bound inside the clip cavity comparable to the reported NADþ
complex of 1a.[5]
1
used as starting material for the studies in buffer. The H NMR
spectra of the molecular clips 1b–d (shown in the Supporting
Information) are slightly concentration-dependent in D2O and in
buffered aqueous solution as well. From this finding, we
extrapolated that these compounds form relatively weak
self-assembled dimers (25 8C: Kdim [Mꢁ1] ꢂ 30 (1b in buffer),
55 ꢃ 8 (1c in D2O), and 140 ꢃ 60 (1d in buffer). Thus, the
molecular clips 1b–d having naphthalene sidewalls largely exist
as monomers in dilute aqueous solution contrary to the
corresponding phosphonate-substituted clip having anthracene
sidewalls that forms a highly stable self-assembled dimer in water
or buffer (25 8C: Kdim [Mꢁ1] ¼ 1.6 ꢀ 105 or 2.3 ꢀ 105).[14]
NMR measurements: host–guest binding
1
Due to the magnetic anisotropy of the clip arene-units H NMR
spectroscopy represents a very sensitive probe for uncovering
the complexation mode of guest molecules. The binding of the
guest molecule inside the clip cavities can be easily detected by
On excitation with ultraviolet light at 280 nm the molecular
clips 1b and 1d show a relatively strong fluorescence band at
340 nm which can be assigned to the emission of the
naphthalene sidewalls (Figure 2).[18] The observation, that this
1
pronounced up-field shifts of the guest signals in the H NMR
spectrum of a host–guest mixture.[3] The binding of the already
mentioned guest molecules by the molecular clips 1b,d
substituted by lithium methanephosphonate or lithium phos-
phate groups at the central benzene spacer-unit were studied in
fluorescence band is quenched by successive addition of NADþ
2
to a solution of 1b or 1d, indicates an electronic interaction of the
clip naphthalene sidewalls with the NADþ nicotinamide ring and/
or adenine unit. This finding was used to determine the binding
Table 1. Maximum complexation-induced 1H NMR shifts (Ddmax) determined for the binding of NADþ 2 with the clips 1b and 1d in
1
buffered aqueous solution at pH ¼ 7.2 or with 1b and 1c in D2O by H NMR titration at 25 8C
Ddmax (ppm) (guest)
Host
Solvent
N-2-H
N-4-H
N-5-H
N-6-H
N-10-H
A-2-H
A-8-H
A-10-H
1b
1d
1d
Buffer
Buffer
Buffera
1.20
1.22
1.41
2.88
2.76
3.23
3.17
3.18
3.69
1.54
1.51
1.75
0.28
0.22
0.32
0.52
0.41
0.57
1.61
0.90
1.68
0.59
0.51
0.60
1b
1c
D2O
D2O
0.39
0.49
0.73
0.90
0.78
—
0.44
0.69
0.32
0.23
—
0.45
0.34
—
0.23
0.51
a Titration with constant guest concentration.
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

J. POLKOWSKA ET AL.
Figure 2. Changes in the fluorescence spectra of a solution of clip 1b (top: 2.88 ꢀ 10ꢁ5 M) and 1d (bottom: 2.63 ꢀ 10ꢁ5 M), respectively upon titration
with NADþ 2 in buffered aqueous solution at pH ¼ 7.6. The insets show the titration curves obtained by plotting the differences in the emission intensities
DI ¼ I0 ꢁ I at 340 nm as function of the equivalents of NADþ 2. The solid lines show the fitting based on the formation of 1:1 complexes
constants of the complex formation between clips 1b,d and
NADþ 2 by fluorometric titration independently to the NMR
measurements. The binding constant determined for the
formation of complex 2ꢄ1b by the NMR titration experiment is
in good agreement with that obtained by fluorometric titration,
whereas in the case of complex 2ꢄ1d the Ka value derived from
the fluorometric titration is significantly larger than that
determined from NMR measurement (Table 2). The difference
between these two Ka values may be explained to be a result of
the weak self-association of 1d which is only important at higher
2ꢄ1b,d, a remarkable splitting is observed for the 1H NMR signals
of the naphthalene protons of 1b,d, which are equivalent in the
free clips (Figure 3(a)). In the complexed clips, the naphthalene
protons Ha and Hf show two separate singlets and the protons
Hb–He an ABCD spectrum instead of one singlet and an A2B2
spectrum, respectively, as found in the free clips. The spectrum of
the protons Hb–He calculated for an ABCD spin system is in good
accord with the experimental data. Evidently, complex dis-
sociation and association as well as conformational equilibration
between the different complex structures proceed rapidly with
respect to the ‘‘NMR time scale.’’[16,17] These dynamic processes
lead to an averaging of the NMR signals of the complex and make
the corresponding protons at the two naphthalene sidewalls of
complexed 1d equivalent to each other, whereas the front- and
backside protons of each naphthalene sidewall remain none-
1
concentrations of 1d as used in the H NMR titration.
Interesting information about the dynamic nature of these
1
host–guest complexes comes from the chemical H NMR shift
nonequivalence in the clip’s highly symmetrical naphthalene
sidewalls induced by the binding of NADþ 2. In the complexes
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790

A STUDY OF THE PH-DEPENDENT BINDING MODE OF NADþ
Table 2. Association constants, Ka, and free association enthalpies, DG, determined for the binding of NADþ 2 with clips 1b and 1d
1
in buffered aqueous solution at pH ¼ 7.2 or 7.6 by HNMR or fluorometric titration (FT) at 25 8C
Host
Solvent
Ka (Mꢁ1
)
DG (kcal/mol)
Method
1b
Buffer (pH ¼ 7.2)
Buffer (pH ¼ 7.6)
4000 ꢃ 330
4900 ꢃ 200
ꢁ4.9
ꢁ5.0
1H NMR
FT
1d
Buffer (pH ¼ 7.2)
4800 ꢃ 1300
ꢁ5.0
1H NMR
FT
Buffer (pH ¼ 7.6)
7100 ꢃ 250
ꢁ5.3
quivalent. This finding can be explained with a chirality transfer of
the chiral guest NADþ 2 to the achiral clip 1d as depicted in
Figure 3(b) for one complex structure including the nicotinamide
ring inside the clip cavity.
found for the complex formation between NMNA 3 as guest
molecule and the molecular clips 1b,d as host molecule (Table 3).
The Ddmax values are similar to the values for the corresponding
protons of the nicotinamide part in the host–guest complexes
with NADþ 2 (Table 1). These data suggest the nicotinamide
ring of NMNA 3 to be positioned inside the cavity of clip 1b or
1d comparable to the recently reported findings for clip 1a
and NMNA 3.[5] This assumption is confirmed by ab initio
Host–guest complexes with NMNA 3, NMN 4, AMP 5, and
NADH 6
1
1
Large binding constants, Ka, and Ddmax values for the H NMR
signals assigned to the protons at the nicotinamide ring were
calculations of the H NMR chemical shifts of the guest protons
(vide infra).
Figure 3. (a) Comparison of the 1H NMR ABCD spin system calculated by WinDaisy (Bruker, top) with the experimental 1H NMR spectra of the protons
Ha ꢁ Hf of clip 1d in the NADþ complex 2ꢄ1d (middle) and in pure clip 1d (bottom) in D2O. (b) A formal 1808 rotation around the C2 axis through the
central benzene spacer-unit of the clip molecule leads to a pair-wise exchange of the host protons (e.g. Hc and Hd) at the naphthalene sidewall marked in
blue with those of the other onꢁe marked in red as it is shown for one complex structure including the nicotinamide ring inside the clip cavity (the structure
for complex 2ꢄ1 (R ¼ OP(OH)O2 ) was calculated by force field (MacroModel, Monte-Carlo conformer search, AMBER /H2O)).[16] Similar conclusions can be
*
drawn for other complex structures, e.g. for those including the adenine system inside the clip cavity. In any case, the corresponding host protons at one
and the same naphthalene sidewall remain nonequivalent
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

J. POLKOWSKA ET AL.
1
Table 3. Association constants, Ka, free association enthalpies, DG, and maximum complexation-induced H NMR shifts, Ddmax
determined for binding of NMNA 3 with the clips 1b,d in buffered aqueous solution at pH ¼ 7.2 by HNMR titration at 25 8C
,
1
Ddmax(ppm) (guest)
Host
Ka (Mꢁ1
)
DG (kcal/mol)
2-H
4-H
5-H
6-H
N–CH3
1b
1d
11 200 ꢃ 1100
33 800 ꢃ 1700
ꢁ5.5
ꢁ6.2
1.61
1.81
3.47
3.69
3.07
3.32
2.25
2.62
0.99
0.99
The comparison of NADþ 2 with the nucleotides NMN 4 and
AMP 5 (the two fragments formally obtained by hydrolysis of the
diphosphate linkage of NADþ 2) allows us to determine the
specific contributions of the nicotinamide and adenine unit
toward the structure and stability of the NADþ complexes. This is
especially important in light of the question, if both heterocycles
are accommodated inside the clip cavity as assumed from Ddmax
values observed for the NADþ complexes. Both nucleotides form
complexes with the clips 1b and 1d in aqueous buffer. All
complexes 4ꢄ1b,d and 5ꢄ1b,d display large complexation-
induced 1H NMR shifts, Ddmax, of the guest protons similar to
those found for the corresponding protons in the NADþ
complexes 2ꢄ1b and 2ꢄ1d (Tables 1 and 4). Inclusion of each
heterocycle in the clip cavity strongly supports the above
postulate that also in the NADþ complexes both ring systems are
positioned inside the clip cavity. Our quantum-chemical studies
with the electron-rich character of the enamine structure of the
dihydronicotinamide ring in NADH as opposed to the extremely
electron-poor pyridinium cation of NADþ, as illustrated in
electron potential surface (EPS) calculations.[5] A molecular clip
such as 1b or 1d is thus able to distinguish between both
oxidation states of NADþ/NADH. This is important for compe-
tition experiments with enzymes – the clip is not prone to
undergo product inhibition with NADH, the product of enzymatic
oxidation.
Another interesting observation comes from the comparison
of the binding constants determined for the formation of the
host–guest complexes of NADþ 2, NMN 4, AMP 5, and NADH 6
with clips 1b,d: the constants determined for the complexes of
NMN 4, AMP 5, and NADH 6 are significantly smaller than those
found for the NADþ complexes. This finding indicates that the
multiple aromatic pꢁp and cationꢁp host–guest interactions in
the NADþ complexes have a synergetic effect on the complex
stability.
1
of complexation induced H NMR shifts (described later in this
work) further show that the structure, where the nicotinamide
part of the NADþ is positioned inside the cavity yields chemical
shifts in better agreement with the experimental results, so that it
seems to be preferred.
NADH 6, the product of enzymatic NADþ reduction, forms
complexes with 1b and 1d in buffered aqueous solution, too. The
large Ddmax values observed for the adenine guest protons and
the negligibly small values found for the protons assigned to the
dihydronicotinamide ring of 6 allow the conclusion that only the
adenine ring is included into the clip cavity. This fact coincides
1
Quantum-chemical calculations of H NMR shifts and
binding energies
In order to gain further insight into possible binding modes of the
host–guest complexes studied in this work, all host–guest
complexes with NADþ, NMNA, NMN, AMP, and NADH as guest
molecules have been studied by quantum-chemical ab initio
calculations of 1H NMR shieldings based on force-field optimized
1
Table 4. Association constants, Ka, free association enthalpies, DG, and maximum complexation-induced H NMR shifts, Ddmax
,
determined for binding of NMN 4, AMP 5, and NADH 6 with 1b and 1d in buffered aqueous solution at pH ¼ 7.2 by 1H NMR titration
at 25 8C
Ddmax (guest) (ppm)
Host
NMN
Ka (Mꢁ1
)
DG (kcal/mol)
2-H
4-H
5-H
6-H
10-H
40-H
1b
1d
550 ꢃ 45
ꢁ3.7
ꢁ4.2
1.28
1.50
2.88
3.20
3.49
4.04
1.75
1.97
0.62
0.61
0.07
0.09
1250 ꢃ 70
AMP
2-H
8-H
10-H
30-H
1b
1d
910 ꢃ 85
680 ꢃ 60
ꢁ4.0
ꢁ3.9
0.38
0.74
1.51
3.14
0.27
0.48
0.03
0.08
NADH
N-2-H
N-4-H
N-10-H
A-2-H
A-8-H
A-10-H
1b
1d
800 ꢃ 60
400 ꢃ 25
ꢁ4.0
ꢁ3.5
0.22
0.12
-
0.2/0.3
0.24
0.10
1.26
0.90
0.49
3.14
0.29
0.53
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790

A STUDY OF THE PH-DEPENDENT BINDING MODE OF NADþ
structures (refer to ‘‘Computational Details’’ section and Support-
ing Information). We focus here on the complexation-induced 1H
NMR chemical shifts for the guest molecule calculated as the
difference between the shieldings of the free and bound guest.
Due to the flexibility of the present host–guest complexes a
multitude of possible structures need to be studied, so that for
each system various phosphate- or phosphonate-substituted
structures have been calculated with a minimum of six structures
for each host–guest complex. For the free NADþ and NADH
guests, three possible structures were investigated, whereas for
NMNA, NMN, and AMP one free guest structure was considered.
Since all the ab initio calculations have been performed in the gas
phase, we focus in the following discussion only on hydrogen
atoms located at aromatic or sugar rings not prone to hydrogen
bonding. For each host–guest complex the proton chemical
shieldings of at least one of the investigated structures agree
within 1.4 ppm with the experimental results obtained in
buffered aqueous solution. This agreement can be considered
sufficient to assign the experimentally observed, large complex-
ation-induced shieldings of up to 4 ppm, so that the structure of
the host–guest complexes can be understood. Here, the error
bars of the calculations arising due to uncertainties within the
force-field optimized structures, the chosen quantum-chemical
approximation for calculating NMR chemical shifts, and the
neglect of solvent effects can be estimated to be typically in the
order of 1 ppm for the considered protons.[5,21]
A comparison between the experimental and calculated Ddmax
data for the host–guest complexes of phosphonate and
phosphate clip 1b and 1d with all guest molecules 2–6 is
shown in Table 5. From the calculations only the data are listed
which give the best fit with the experimental values. The
observed differences between the experimental and calculated
data result from the difficulties mentioned above and the fact
that each host–guest complex certainly consists of several
conformers which are in a dynamic equilibrium and contribute to
1
its overall structure so that the experimental chemical H NMR
shifts are averaged values. More detailed data are presented in
the Supporting Information, while we focus in the following on
the most important aspects for the NADþ host–guest complex (as
shown in Figure 4).
The two fundamentally different possibilities of binding NADþ
either by the adenine or the nicotinamide unit can be
distinguished by comparing quantum-chemically computed
and experimental NMR data: here, we considered six structures
where nicotinamide is bound inside the clip cavity and three
structures with the adenine part inside the cavity (of these
nine structures, one nicotinamide and one adenine bound
structure were taken from Reference [5]; the other seven
structures were calculated in this work, refer to Supporting
Information for details).
The nicotinamide-bound structures clearly show a better
agreement with the experiment than the ones with adenine
1
Table 5. Comparison of the experimental complexation-induced H NMR shifts, Ddmax, of the guest protons in buffered aqueous
solution with those calculated by ab initio methods
Complex
N-2-H
N-4-H
N-5-H
N-6-H
N-10-H
A-2-H
A-8-H
A-10-H
NADþ 2ꢄ1b
1.2
2.0
1.6
2.9
3.7
2.2
3.2
3.3
0.7
1.5
1.8
0.8
0.3
1.0
0.9
0.5
0.5
2.6
1.6
1.0
1.4
0.6
0.4
0.5
NAD-Nica
NAD-Adea
NADþ 2ꢄ1d
NAD0-Nica
NAD0-Adea
1.2
2.2
1.5
2.8
3.8
1.4
3.2
3.6
0.3
1.5
1.5
ꢁ0.1
2.3
0.2
1.6
0.6
0.4
0.9
2.9
0.9
0.4
1.1
0.5
1.0
0.9
NMNA 3ꢄ1b
1.6
1.7
3.5
3.4
3.1
3.3
1.0
1.5
—
—
—
—
—
—
NMNA6a
3.0
NMNA 3ꢄ1d
1.8
2.1
3.7
3.6
3.3
3.0
2.6
1.4
1.0
0.9
—
—
—
—
—
—
NMNA2a
NMN 4ꢄ1b
1.3
1.6
2.9
3.1
3.5
3.0
1.8
1.8
0.6
0.6
—
—
—
—
—
—
NMN6a
NMN 4ꢄ1d
1.5
1.6
3.2
5.2
4.0
2.6
2.0
2.5
0.6
1.1
—
—
—
—
—
—
NMN1a
AMP 5ꢄ1b
—
—
—
—
—
—
—
—
—
—
0.4
2.7
1.5
1.5
0.3
0.8
AMP9a
AMP 5ꢄ1d
—
—
—
—
—
—
—
—
—
—
0.7
0.9
3.1
2.7
0.5
1.9
AMP2a
NADH 6ꢄ1b
0.2
ꢁ
ꢁ
ꢁ
0.2
1.3
2.4
0.5
1.0
0.3
1.0
NADH8a
ꢁ0.5
0.1
ꢁ0.7
0.2/0.3
ꢁ1.1
—
ꢁ1.5
—
ꢁ0.2
0.1
NADH 6ꢄ1d
0.9
1.0
3.1
4.4
0.5
1.6
NADH3a
1.6
ꢁ0.5
ꢁ0.5
0.0
0.5
a Calculated structure (as shown in Figure 4 or Supporting Information: Table QC-1 – QC-9; NAD-Nic ¼ NADþ5, NAD-Ade ¼ NADþ9,
NAD0-Nic ¼ NADþ3, NAD0-Ade ¼ NADþ7, compared to G3; NADH8 is compared to free guest G2 and NADH3 to G3).
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

J. POLKOWSKA ET AL.
The good agreement of the quantum-chemical and exper-
imental results for the nicotinamide-bound complex provides
strong evidence that it is the most prominent structure in the
complexation process, while complexation of the adenine part of
NADþ is less dominant. At the same stage it is important to
stress, that the flexibility of the host–guest complexes is rather
large and a dynamic process between various binding motifs is
expected.
Binding energies
Besides the results derived above by comparing computed and
experimental complexation-induced chemical shieldings, there
are also energetic indications for the preferred binding via the
nicotinamide unit of NADþ. While we have published already
interaction energies for the example of two possible structures
denoted in Reference [5] as NAD-1 and NAD-2 without
counterpoise correction, we provide in our present work also
interaction energies computed at the most reliable RI-MP2/SVP
level with counterpoise correction (as shown in Table 6). The
counterpoise corrected data indicate that the interaction energy
within the nicotinamide bound structure NAD-1 is stronger by 3
to 27 kcal/mol as compared to the one for adenine binding
(NAD-2) depending on the individual charges of the complex
components. As expected, the counterpoise correction slightly
reduces the interaction energies, however, the relative energies
of nicotinamide versus adenine binding are only influenced by
4 kcal/mol at most. A similar behavior of basis-set superposition
effects is also found for the host–guest interaction within a
molecular tweezer binding dicyano-benzene: the total inter-
action energy as computed at the RI-MP2/SVP level is reduced
from 39 to 25 kcal/mol, however, the additivity of interaction
energies is preserved upon counterpoise correction; one part of
the tweezer binds the guest by 8kcal/mol and the other two parts
by 7 kcal/mol which adds up to 22 kcal/mol. This means that 88%
of the binding energy results from the aromatic parts of the clip,
while without counterpoise correction it is fairly similar with a
contribution of 95%.
*
Figure 4. AMBER /H2O structures NAD-Nic (top) and NAD-Ade (bottom)
of complex 2ꢄ1 (R ¼ OP(Me)Oꢁ2 ) and the complexation-induced 1H NMR
shifts (Ddmax) [ppm] computed at the GIAO-HF/SVP level as differences of
the chemical 1H NMR shifts (d) for free and complexed guest protons. The
listed data are the calculated Ddmax values in ppm. The displayed
structures NAD-Nic (nicotinamide binding) and NAD-Ade (adenine bind-
ing) are the ones for which the calculated Ddmax values show the smallest
deviations to the experimental data (in buffered aqueous solution).
Structure NAD-Ade was taken from Reference [5].
Table 6. Host–guest interaction energies (Eint) in kcal/mol
within NAD-1 and NAD-2 structures[5] as obtained at the
RI-MP2/SVP level both without and with counterpoise
corrections (cp-corr) For NAD-2 the differences of the inter-
action energies (DE ¼ Eint (NAD-1) ꢁ Eint (NAD-2)) are given
Difference
(NAD-1 ꢁ NAD-2)
Charge
NAD-1
inside the clips’ cavity: for the nicotinamide-bound complexes
the best agreement is observed for the NAD-Nic structure
displayed in Figure 4 with maximum deviations for the calculated
1H NMR shifts of nicotinamide of 0.8 ppm, while the chemical
shifts of its adenine unit outside the cavity agree within 0.6 ppm.
Complex-induced shieldings of other nicotinamide-bound struc-
tures deviate stronger by 1.4–3.9 ppm for their nicotinamide unit
and 0.5–1.1 ppm for their adenine unit.
In contrast, the alternative binding modes via adenine show
maximum deviations of 2.1–2.6 ppm in comparison to the
experiment for their complexed unit, while deviations of
2.5–3.7 ppm are computed for their nicotinamide part outside
the cavity.
Clip NADþ
Eint
Eint (cp-corr) DEint DEint (cp-corr)
ꢁ2
ꢁ2
0
þ1a
ꢁ152
ꢁ46
ꢁ34
ꢁ79
ꢁ78
ꢁ171
ꢁ136
ꢁ30
ꢁ18
ꢁ48
ꢁ47
ꢁ139
—
ꢁ9
—
ꢁ13
ꢁ3
0a
0a
þ1
0
0
þ1
ꢁ8
ꢁ8
0
ꢁ7
ꢁ8
ꢁ2
þ1
ꢁ27
ꢁ27
a Only the part of the guest molecule bound inside the clip is
considered.
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790

A STUDY OF THE PH-DEPENDENT BINDING MODE OF NADþ
R
R
O
4
a: OP(Me)O2– (nBu)4N+
b: OP(Me)O2– Li+
c: OP(OH)O2 Li+
d: OPO32– 2Li+
–
5
6
NH2
R
2
N
I
1
CH3
e: OPO3H2
NMNA 3
N
4
A
O
NH2
N
5
6
N
N
NH2
8
2
2
N
N
O
P
O
P
O
O
O
O
OH
OH
1'
1'
2'
3'
O
O
O
OH OH
NAD+ 2
O
NH2
N
4
N
N
5
6
NH2
N
8
2
2
N
O
P
O
P
O
O
O
O
O
O
1'
1'
2'
OH
2'
O
4' 3'
O
3'
4'
OH
OH OH
NMN 4
AMP 5
N
A
O
NH2
4
5
6
N
N
NH2
N
8
2
2
N
N
O
O
P
O
P
O
O
O
OH
OH
1'
1'
2'
3'
O
O
O
OH OH
NADH 6
Scheme 1. Structures of molecular clips 1a–e as well as NADþ 2, NMNA 3, NMN 4, AMP 5, NADH 6 (N ¼ nicotinamide; A ¼ adenine) with their respective
numbering schemes
of 1 mol equivalent of clip 1a, 1b, or 1c and, hence, cannot be
accommodated inside the clip cavity. In the formal 1:1 complexes
CONCLUSION
NADþ 2 forms host–guest complexes with the molecular clips
1b–d displaying large complexation-induced 1H NMR shifts,
Ddmax, of the guest signals assigned to the nicotinamide as well as
adenine protons in buffered aqueous solution at pH ¼ 7.2. In
contrast, NMR shifts of the corresponding complexes remain
small in pure water. We conclude that in buffer at neutral pH
value, NADþ 2 forms host–guest complexes with clips 1b–d,
where either the nicotinamide or the adenine part can be bound
inside the clip cavity, whereas in the complexes dissolved in pure
water under slightly acidic conditions both heterocyclic units are
located outside the clip cavity. According to ITC measurements
observed by the NMR dilution experiments, we assume that clip
molecules stick to the outer surface of the NADþ aggregate.
Interestingly, this assumption is supported by the finding that in
pure water the complexation-induced 1H NMR shifts of the NADþ
guest signals increase by addition of excess clip 1b or 1c
approaching the Ddmax values observed in buffered aqueous
solution. The excess of host, evidently, affects the dissociation of
the NADþ aggregate by clipping one of the heterocyclic units.
The true structural switch between aggregated and monomeric
NADþ is, however, triggered by deprotonation. In water under
slightly acidic conditions, the NADþ molecule exists in its
zwitterionic form presumably protonated at nitrogen atom N-1 of
the adenine unit.[12] Thus, it forms the higher aggregate detected
the acidic aqueous solution contains aggregated NADþ
2
molecules which evidently do not dissociate in the presence
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

J. POLKOWSKA ET AL.
by the ITC measurements most likely resulting from either
hydrophobic or electrostatic interactions.[23–26] In buffered
solution at pH ¼ 7.2, however, the overall neutral NADþ
molecules are deprotonated and, hence, become negatively
charged. The negative overall charge destabilizes existing
aggregates leading to efficient monomerization of the cofactor.
Monomeric NADþ forms clean (1:1) complexes with the clips
1b and 1d including either the nicotinamide or the adenine unit
inside the clip cavity. If both heterocyclic units are included, the
clip has to shuttle in the (1:1) complex from one to the other unit
rapidly on the ‘‘NMR time scale’’ in order to explain the observed
complexation-induced 1H NMR up-field shifts of the signals
assigned to the nicotinamide and adenine guest protons. As an
additional rapid process, a formal 1808 rotation of the clip
molecule around the guest unit was detected by the complex-
ation-induced 1H NMR shifts of the host protons resulting from a
chirality transfer of the chiral guest to the achiral host (Figure 3).
These two dynamic processes must occur via mutual complex
dissociation and association and not by ‘‘intramolecular’’
conformational isomerizations, as they were found for the guest
rotation and shuttle processes inside the clip or tweezer cavity of
other host–guest systems.[16,17] Our conclusion, that both units of
NADþ 2 are bound inside the clip cavity, is supported by the
observation that NMNA 3 and the nucleotides NMN 4 and AMP 5
form complexes with the clips 1b and 1d showing large Ddmax
values for the guest 1H NMR signals assigned to the nicotinamide
or adenine protons comparable to those found for the
corresponding protons of the NADþ complexes. Intriguingly,
quantum-chemically calculated complexation-induced 1H NMR
shifts for the nicotinamide bound NADþ complex show a better
agreement with the experimentally observed shieldings than the
ones calculated for the adenine bound NADþ complex.
Importantly, complexes of NMN 4, AMP 5, and NADH 6 are
significantly less stable than those of NADþ 2, indicating that
multiple aromatic pꢁp and cationꢁp host–guest interactions
occurring in the NADþ host–guest complexes have a synergetic
effect on the complex stability. The clips’ (1b,d) fluorescence
emission is almost entirely quenched by complex formation with
NADþ – an excellent starting point for the development of
chemical sensors.
The inclusion of NADþ and other enzyme cofactors at their
active sites by molecular clips is a prerequisite for the systematic
investigation of their effects on enzymatic cofactor-mediated
processes. Applying the optimum conditions for efficient NADþ
inclusion inside the clip cavity detailed in this work, we recently
proceeded to experiments with cofactor-dependent enzymes.
Indeed, the enzymatic oxidation of alcohol catalyzed by AD[10]
using NADþ as cofactor is inhibited by the addition of clip 1b or
1d. Detailed mechanistic investigations point to a typical
behavior of both clips as competitive inhibitors (Lineweaver–Burk
plots, inhibition reversal by cofactor addition).[10] We conclude
that also under physiological conditions the molecular clips are
able to capture NADþ and deplete the cofactor concentration
inside the Rossman fold.
1H-NMR titration experiments: DRX 500. The undeuterated
amount of the solvent was used as an internal standard. The
1H and 13C NMR signals were assigned by the 2D experiments
mentioned above. Positions of the protons of the methylene
bridges are indicated by the letters i (innen, toward the center of
the molecule) and a (aussen, away from the center of the
molecule). MS: Fison Instruments VG ProSpec 3000 (70 eV). All
melting points are uncorrected. Column chromatography: silica
gel 0.063–0.2 mm. All solvents were distilled prior to use. The
aqueous phosphate buffer at pH ¼ 7.2 (used for the 1H NMR
titration experiments) was prepared by dissolving NaOH
(1.06 mmol) and KH2PO4 (1.33 mmol) in 20 ml of D2O and that
at pH ¼ 7.6 (used for the fluorometric titration experiments) by
the addition of NaOH (10%) to a solution containing Na2HPO4
(100 mM), NaH2PO4 (100 mM), MgCl2 (7 mM), and ethylenedia-
mine tetraacetic acid, EDTA (0,1 mM) until the desired pH value of
7.6 has been achieved.
The molecular clips 1b and 1d were synthesized and
characterized as previously described.[6]
Dilithium-(6a, 8a, 15a, 17a)-6, 8, 15, 17-tetrahydro-6:17,
8:15-dimethanoheptacenyl-7,16-bisphosphate 1c
LiOH ꢅ H2O (9.6 mg, 228.9 mmol) is added to the stirred solution of
phosphoric acid 1e (68.7 mg, 114.9 mmol) in 5 ml of methanol at
room temperature. After stirring of the clear solution for 30 min,
methanol is evaporated in vacuo. The residue is dried several
hours in vacuo to give 70 mg of 1c as beige solid in quantitative
yield. Mp ¼ 223 8C (decomposition); 1H NMR (500 MHz, D2O):
d ¼ 2.41 (d, 2 H, 2J (19-Ha, 19-Hi) ¼ 8.2 Hz, 19-Ha, 20-Ha), 2.69 (d, 2
H, 19-Hi, 20-Hi), 4.73 (s, 4 H, 6-H, 8-H, 15-H, 17-H), 6.87 (m, 4 H, 2-H,
3-H, 11-H, 12-H), 6.97 (m, 4 H, 1-H, 4-H, 10-H 13-H), 7.30 (s, 4 H, 5-H,
9-H, 14-H, 18-H); 13C NMR (125.7 MHz, D2O): d ¼ 48.40 (s, CH2,
C-19, C-20), 64.46 (s, CH, C-6, C-8, C-15, C-17), 120.23 (s, CH, C-5,
C-9, C-14, C-18), 125.32 (s, CH, C-2, C-3, C-11, C-12), 127.66 (s, CH,
C-1, C-4, C-10, C-13), 131.88 (s, C-4a, C-9a, C-13a, C-18a), 141.95 (s,
C-6a, C-7a, C-15a, C-16a), 147.55 (s, C-5a, C-8a, C-14a, C-17a) the
signal of the carbons C-7 and C-16 is not observed; 31P NMR
(202 MHz, D2O): d ¼ ꢁ3.01 (s, 2 P, OP(O)(OH)OLi).
The association constants Ka and the complexation-induced
chemical 1H NMR shifts, Ddmax, were determined by 1H NMR
titration as described in References [5,6] Host H and guest G are in
equilibrium with the 1:1-complex HG. The association constant Ka
is then defined by Eqn (1). [H]0 and [G]0 are the starting
concentrations of host and guest, respectively.
½HGꢆ
½HGꢆ
ꢀ
ꢁ ꢀ
ꢁ
Ka ¼
¼
(1)
½Hꢆ ꢅ ½Gꢆ
½Hꢆ0 ꢁ ½HGꢆ ꢅ ½Gꢆ0 ꢁ ½HGꢆ
The observed chemical shift dobs of the guest proton in the 1H
NMR spectrum (Bruker instrument DRX 500, 500 MHz, 25 8C) of a
host and guest mixture is an averaged value between free (d0)
and complexed guest (dHG), in the case of the exchange between
free and complexed guest being fast with respect to the NMR
time scale (Eqn (2)). Combination of Eqns (1) and (2) and the use
of differences in chemical shift (Dd ¼ d0 ꢁ dobs; Ddmax ¼ d0 ꢁ dHG
)
EXPERIMENTAL SECTION
leads to Eqns (3) and (4), respectively.
General experimental details
½Gꢆ
½HGꢆ
1
dobs
¼
ꢅ d0 þ
ꢅ dHG
(2)
IR: Bio-Rad FTS 135. UV: JþM Tidas FG Cosytec RS 422. H-NMR,
½Gꢆ þ ½HGꢆ
½Gꢆ þ ½HGꢆ
13C-NMR, DEPT H,H-COSY, C,H-COSY, NOESY, HMQC, HMBC,
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790

A STUDY OF THE PH-DEPENDENT BINDING MODE OF NADþ
ꢂ
ꢃ
280 nm and the fluorescence intensity was monitored at 340 nm.
The association constants, Ka, were determined from the fitting of
the data pairs – DIobs and [G]0/[H]0 – to the binding isotherm
given in Eqn (6):
Ddmax
½Gꢆ0
1
2
1
Dd ¼
ꢅ
½Hꢆ0 þ ½Gꢆ0 þ
!
Ka
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
(3)
ꢂ
ꢃ
2ꢁ½Hꢆ0 ꢅ ½Gꢆ0
1
4
1
ꢁ
ꢅ
½Hꢆ0 þ ½Gꢆ0 þ
Ka
DI ½Hꢆ0 þ ½Gꢆ0 þ 1=Ka
I ¼ I0 þ
ꢄ
ꢅ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
½Gꢆ0
2
K
2K2
½Hꢆ0Ka
2K
K2
Ddmax
2
Dd ¼
K þ 1 þ
ꢁ
K2 þ
ꢁ 2K þ
þ
þ 1
½Hꢆ0Ka
2
½Hꢆ0Ka2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
½Hꢆ0Ka
(6)
!
2
ð½Hꢆ0 þ ½Gꢆ0 þ 1=KaÞ
ꢁ ½Hꢆ0½Gꢆ0
½Hꢆ0
ꢁ
with K ¼
½Gꢆ0
4
(4)
Two types of titration experiments were performed: (i) in the
case of Ka ꢂ 4000 Mꢁ1, the total guest concentration [G]0 was
kept constant, whereas the total host concentration [H]0 was
varied (Eqn (3)); (ii) in the case of Ka ꢇ 4000 Mꢁ1, dilution titrations
were performed by the use of solutions containing equimolar
amounts of host and guest with concentrations in the range
between 0.5 and 4 mM (Eqn (4)). In both experiments, Dd was
determined as the difference of the chemical shifts of the proton
of pure guest (d0) and of the guest in the various host–guest
mixtures (dobs). The dependence of dobs on the host concen-
trations afforded the data pairs – Dd and [H]0. Fitting of these data
to the 1:1 binding isotherm (Eqns (3) and (4), respectively) by
iterative methods using the computer program TableCurve 2D,
version 5.0 delivered the parameters Ka and Ddmax. In the case of
mixed (1:1) and (1:2) or (2:1) complex stoichiometries the data
pairs – dobs and [H]0 – were used for the fitting of the binding
isotherm in the computer program HOSTEST version 5.60.[15] Since
the guest molecules (used in this work) possess more than one
kind of nonequivalent protons, Ka and Ddmax were determined for
the guest protons which can be unambiguously assigned over the
total range of concentration and show substantial complex-
ation-induced shifts, Dd. In these cases the average of the Ka values
(determined independently) is given in Tables 2–4. If the guest
proton signals cannot be unambiguously assigned at each
concentration because they are broadened or superimposed by
other NMR signals, the Ddmax values of these guest protons were
calculated by the use of Eqn (5). The experimental error was
estimated to be small for the Ddmax values (<5%), but large for the
binding constants Ka (ꢂ20%), particularly in the case of the highly
stable complexes for which these data have to be determined by
NMR dilution titrations as described above.
Isothermal titration calorimetry (ITC) measurements
ITC measurements were carried out by the use of a Microcal
Calorimeter at 25.0 8C (VP-ITC Microcalorimeter; MicroCal).
Computational details
The ab initio calculations have been carried out with a
development version of the Q-Chem program package.[27]
NMR chemical shifts were calculated using the GIAO-HF
Scheme[28–30] with the SVP basis set,[31] employing the
linear-scaling
density matrix-based
GIAO-HF method
(D-GIAO-HF) described in References [21,32,33]. Relative chemical
shifts were obtained by referring to a TMS molecule calculated at
the same level of theory. Structures were optimized by force-field
(Monte-Carlo conformer search employing structures calculated
[19,20]
*
by AMBER /H2O, MacroModel),
which typically provides a
good starting point for the calculation of NMR chemical
shieldings in the present systems, as it was demonstrated
in previous work on a simple model system.[34] All binding
energies were calculated using the RI-MP2 method (refer
to References [35,36] and references therein) with the SVP basis
set.
Acknowledgements
This work was supported by the Deutsche Forschungsge-
meinschaft (DFG). The authors thank Heinz Bandmann and
Dr. Torsten Schaller for NMR measurements, Klaus Kowski
for ITC measurements, Willi Sicking for assistance with the
HOSTEST calculations, and Professor Craig Wilcox for providing
us access to the HOSTEST program. C.O. acknowledges
financial support by an Emmy Noether research grant and the
‘‘Graduiertenkolleg’’ GK 441 ‘‘Chemie in Interphasen’’ of the
Dd1;max
Ddn;max ¼ Ddn
(5)
Dd1
¨
DFG. J.Z. thanks the ‘‘Landesgraduiertenforderung’’ (LGFG) for a
graduate fellowship.
Job-plot analysis
Equimolar solutions of host and guest compound (each
approximately 10 mmol/10 ml) were prepared and mixed in
various portions, so that the sum of the portions is constant for
each mixture. 1H NMR spectra of the mixtures were recorded, and
the chemical shifts were analyzed by Job’s method modified for
NMR spectroscopy.
REFERENCES
[1] D. Voet, J. G. Voet, Biochemistry, 3rd edn, Wiley, New York, 2005.
[2] A. Fersht, Structure and Mechanism in Protein Science, A Guide to
Enzyme Catalysis and Protein Folding, W.H. Freeman and Company,
New York, 2000.
[3] Review: F.-G. Kla¨rner, B. Kahlert, Acc. Chem. Res. 2003, 36, 919–
932.
[4] C. Jasper, T. Schrader, J. Panitzky, F.-G. Kla¨rner, Angew. Chem. Int. Ed.
2002, 41, 1355–1358.
[5] M. Fokkens, C. Jasper, T. Schrader, F. Koziol, C. Ochsenfeld, J. Polk-
owska, M. Lobert, B. Kahlert, F.-G. Kla¨rner, Chem. Eur. J. 2005, 11,
477–494. In this study tetra-n-butylammonium cations were used as
counter-ions for the methanephosphonate groups in clip 1a in order
to make this molecule soluble in a variety of solvents from water to
Fluorescence measurements
The fluorescence spectra were measured in Jasco spectro-
fluorometer FP-6500 at 25 8C. The fluorometric titration with
NADþ 2 as guest molecule was conducted by keeping the
concentration of clip 1b or 1c as host molecule constant and
varying of the NADþ concentration. Each sample was excited at
J. Phys. Org. Chem. 2009, 22 779–790
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

J. POLKOWSKA ET AL.
chloroform. An independent study with the lithium methanepho-
sphonate clip 1b as host molecule and tetra-n-butylammonium
bromide as guest molecule showed that 1b forms a (1:1) complex
with the tetra-n-butylammonium cation (Ka ¼ 3800 Mꢁ1).
[6] T. Schrader, M. Fokkens, F.-G. Kla¨rner, J. Polkowska, F. Bastkowski, J.
Org. Chem. 2005, 70, 10227–10237.
[7] Three earlier reports have appeared on molecular recognition of the
NADþ bisphosphate bridge: F. P. Schmidtchen, Chem. Ber. 1981, 114,
597–607.
[17] V. Balzani, H. Bandmann, P. Ceroni, C. Giansante, U. Hahn, F.-G. Kla¨rner,
¨
U. Mu¨ller, W. M. Mu¨ller, C. Verhaelen, V. Vicinelli, F. Vogtle, J. Am. Chem.
Soc. 2006, 128, 637–648.
[18] I. B. Berlman, Handbook of Fluorescence Spectra of Aromatic Molecules,
2nd edn, Academic Press, New York, 1971.
[19] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton, C.
Caufield, G. Chang, T. Hendrickson, W. C. Still, J. Comput. Chem. 1990,
11, 440–467.
¨
[20] Macromodel, v. 7.1, Schrodinger, 1500 SW First Ave., Ste. 1180,
[8] H. Fenniri, M. W. Hosseini, J. M. Lehn, Helv. Chim. Acta 1997, 80,
786–803.
[9] A. Domenech, E. Garcia-Espana, J. A. Ramirez, B. Celda, M. C. Martinez,
D. Monleon, R. Tejero, A. Bencini, A. Bianchi, J. Chem. Soc. Perkin Trans.
2 1999, 23–32.
Portland, OR 97201.
[21] C. Ochsenfeld, J. Kussmann, F. Koziol, Angew. Chem. Int. Ed. 2004, 43,
4485–4489.
[22] C. Ochsenfeld, F. Koziol, S. P. Brown, T. Schaller, U. P. Seelbach, F.-G.
Kla¨rner, Solid State Nucl. Magn. Reson. 2002, 22, 128–153.
[23] Review: E. A. Meyer, R. K. Castellano, F. Diederich, Angew. Chem. Int.
Ed. 2003, 42, 1210–1250.
[10] P. Talbiersky, F. Bastkowski, F.-G. Kla¨rner, T. Schrader, J. Am. Chem. Soc.
2008, 130, 9824–9828.
[11] M. Kirsch, P. Talbiersky, J. Polkowska, F. Bastkowski, T. Schaller, H. de
Groot, F.-G. Kla¨rner, T. Schrader, in press, Angew. Chem.
[12] C. Nadolny, G. Zundel, J. Mol. Struct. 1996, 385, 81–87.
[13] It is worth mentioning that the 1H NMR spectra of NADþ in D2O are
not concentration-dependent and, hence, do not give any hint on its
self-association detected by ITC.
[24] Cation–p interaction: J. P. Gallivan, D. A. Dougherty, J. Am. Chem. Soc.
2000, 122, 870–874.
[25] Hydrophobic effect: C. Tanford, The Hydrophobic Effect, Wiley, New
York, 1980.
[26] Review: R. Breslow, Acc. Chem. Res. 1991, 24, 159–170.
[27] Development version of the Q-Chem program package (http://
www.q-chem.com).
[28] R. Ditchfield, Mol. Phys. 1974, 27, 789–807.
[29] K. Wolinski, J. F. Hinton, P. Pulay, J. Am. Chem. Soc. 1990, 112,
8251–8260.
[14] F.-G. Kla¨rner, B. Kahlert, A. Nellesen, J. Zienau, C. Ochsenfeld, T.
Schrader, J. Am. Chem. Soc. 2006, 128, 4831–4841.
[15] Nonlinear regression analyses of Eqns (4) and (5) were performed by
the use of the program Table Curve 5.0, SPSS Science analogously to
the computer program HOSTEST by Wilcox, C. S. and Glagovich, N.
M., University of Pittsburgh, and the program Associate V1.6, Peter-
son, B. PhD Dissertation, University of California at Los Angeles, 1994.
In order to determine the complex stoichiometry the experimental
were analyzed with the HOSTEST program and Job plots. P. Job, C. R.
Hebd. Seances Acad. Sci. 1925, 180, 925.
[30] M. Ha¨ser, R. Ahlrichs, H. P. Baron, P. Weis, H. Horn, Theor. Chim. Acta
1992, 83, 455–470.
[31] A. Scha¨fer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–2577.
[32] J. Kussmann, C. Ochsenfeld, J. Chem. Phys. 2007, 127, 054103.
[33] M. Beer, C. Ochsenfeld, J. Chem. Phys. 2008, 128, 221102.
[34] J. Zienau, J. Kussmann, F. Koziol, C. Ochsenfeld, Phys. Chem. Chem.
Phys. 2007, 9, 4552–4562.
[16] Conformational equilibration of guest molecules inside the host
cavity could be observed in related host-guest systems by tempera-
[35] F. Weigend, M. Ha¨ser, H. Patzelt, R. Ahlrichs, Chem. Phys. Lett. 1998,
294, 143–152.
1
ture-dependent H NMR spectroscopy: M. Lobert, H. Bandmann, U.
Burkert, U. P. Bu¨chele, V. Podsadlowski, F.-G. Kla¨rner, Chem.-Eur. J.
2006, 12, 1629–1641.
[36] R. A. Kendall, H. A. Fru¨chtl, Theor. Chem. Acc. 1997, 97, 158–
163.
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 779–790