Cytosine-substituted metalloporphyrins: receptors for recognition of nucleotides in ion-selective electrodes.

Article abstract of DOI:10.1039/b400880d

A new series of cytosine-substituted metalloporphyrin conjugates (containing Co(II) and Zn(II) as the coordinated metals) were designed and investigated as nucleotide receptors in PVC-membrane-based ion-selective electrodes under neutral conditions. The potentiometric results indicate that these systems, in particular the Co(II)-containing complex, may be potentially useful sensors for complementary nucleotide substrates in the presence of 10 mol% tridodecylmethylammonium chloride (K(Pot.)(5'-GMP/5'-AMP)= 0.045).

Full text of DOI:10.1039/b400880d

View Article Online / Journal Homepage / Table of Contents for this issue
Cytosine-substituted metalloporphyrins: receptors for recognition
of nucleotides in ion-selective electrodes
Vladimír Král,*a Tatiana V. Shishkanova,a Jonathan L. Sessler*b and Christopher T. Brownb
a Department of Analytical Chemistry, Institute of Chemical Technology, 16628 Prague 6,
Technická 5, Czech Republic. E-mail: vladimir.kral@vscht.cz; Fax: ϩ420 2 2431 0352;
Tel: ϩ420 2 2435 4227
b Department of Chemistry and Biochemistry, Institute for Cellular and Molecular Biology,
The University of Texas at Austin, 1 University Station – A5300, Austin, Texas, 78712-0165,
USA. E-mail: sessler@mail.utexas.edu; Fax: ϩ1 512 471 7550; Tel: ϩ1 512 471 5009
Received (in Pittsburgh, PA, USA) 19th January 2004, Accepted 26th February 2004
First published as an Advance Article on the web 18th March 2004
A new series of cytosine-substituted metalloporphyrin conjugates (containing Co() and Zn() as the coordinated
metals) were designed and investigated as nucleotide receptors in PVC-membrane-based ion-selective electrodes
under neutral conditions. The potentiometric results indicate that these systems, in particular the Co()-containing
complex, may be potentially useful sensors for complementary nucleotide substrates in the presence of 10 mol%
tridodecylmethylammonium chloride (KPot.5Ј-GMP/5Ј-AMP = 0.045).
insights into recognition events that take place at an aqueous–
organic interface.
Introduction
A key step in rendering practical many of the inspirational
results arising from the current explosion in the area of supra-
molecular chemistry involves showing that the selectivity in
recognition demonstrated in solution can be reproduced to
good effect on surfaces, membranes, electrodes and other
supports through which an interface to conventional instru-
mentation may be readily established. One approach we and
others have been pursuing in this regard is focused on the
development of selective, molecular recognition-based ion-
selective electrodes. Here, one of the goals has been and
remains the generation of systems that recognize and sense
mononucleotides. Such species play a critical role in biology
and medicine and thus, not surprisingly, the development of
receptors, carriers, and sensors for nucleotides has attracted the
attention of many researchers working in the supramolecular
field.1–19 From this body of work, it is now well-appreciated that
various combinations of complementary base pairing, directed
multisite hydrogen bonding interactions, specific pi–pi (π–π)
stacking effects, and generalized electrostatic interactions
can be used to design synthetic receptors, just as these various
binding motifs, individually or in concert, are thought to
contribute to the exquisitely sensitive nucleotide recognition
observed in biological systems. Incorporating one or more of
these features into appropriately designed synthetic receptors
could culminate in the production of highly specific potentio-
metric sensors (in particular, ion-selective electrodes, ISEs),
which by virtue of their interfacial nature are potentially well-
suited for use in various “real world” bioanalytical applications.
Whether this goal will be fully realized remains to be seen.
However, its pursuit has stimulated the evolutionary develop-
ment of several elegant nucleotide-targeting ISEs in recent
years.1,3,6,9,11,13 Here, in addition to the ultimate goal of gener-
ating a field-usable working nucleotide sensing system, an
important ancillary motivation for making and studying
various potential carrier-based ISEs, is that their analysis
can provide a means for testing quickly whether a given
supramolecular receptor system displays selectivity for the
targeted analyte(s) under what can be considered model “real
world” conditions.7,13 In particular, the potentiometric methods
involved in the testing of potential ISEs can be used to derive
In recent years, we have devoted considerable effort to the
design of ditopic receptors for the specific recognition of
various targeted nucleotides. To date, we have mostly focused
on the use of systems based on nucleobase-substituted
oligopyrrolic macrocycles, such as sapphyrin and calix[4]-
pyrroles. In the present study, we describe a new set of metallo-
porphyrin–nucleobase conjugates and tests of their use as
potential nucleotide-recognizing ISE sensing elements (Fig. 1).
Results and discussion
In an initial series of experiments, PVC/o-NPOE membrane
electrodes containing no lipophilic additive were tested for their
response to nucleotides. Such membranes, which did, however,
contain the cytosine–metalloporphyrin conjugates, were found
to demonstrate: i) a high potentiometric sensitivity toward
5Ј-CMP (slopes of Ϫ27 and Ϫ34 mV/decade beginning from
10Ϫ3 M for 2 and 3, respectively), ii) a sub-Nernstian response
toward 5Ј-GMP and 5Ј-UMP beginning from 10Ϫ5 M, and iii)
a weak response toward 5Ј-AMP and 5Ј-TMP (Fig. 2).
A comparison of the overall anion responses (∆mV) of the
PVC-membrane electrodes doped with receptors 2 and 3
showed an increasing potentiometric response within the
series: 5Ј-AMP < 5Ј-TMP < 5Ј-GMP < 5Ј-UMP < 5Ј-CMP. It
should also be noted that the potentiometric response of the
membranes based on the Co()-porphyrin 2 was far cleaner
than that of the membranes based on Zn()-porphyrin 3. While
a variety of factors could account for the latter findings, they
could also be providing a “hint” as to which portions of
the receptor are most important in terms of recognizing
nucleotide(s).
In receptors 2 and 3, two types of interactions (or some
combination thereof ) were considered likely to provide the
basis for nucleotide recognition: i) hydrogen-bonding between
the complementary nucleobases present in the receptor and the
targeted nucleotide guest and ii) direct coordination, wherein
the nucleotide guest binds as an axial ligand to the Lewis acidic
metal center of the metalloporphyrin. In an effort to determine
whether the latter mode is likely to be important for nucleotides
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1169

View Article Online
Table 1 Association constants (MϪ1) for the interaction of 3Ј,5Ј-bis(O-triisopropylsilyl) substituted deoxyribonucleosides and zinc octaethyl-
porphyrin, as reported by Ogoshi et al.20
Nucleoside derivatives
Association constanta
Adenosine
Guanosine
Cytidine
Thymidine
600 (100)
150 (50)
80 (50)
—
a In CH2Cl2, at 15 ЊC. Standard deviations are given in parentheses. In the case of thymidine, an appreciable absorption change was not observed.
Data taken from ref. 20.
Fig. 1 Compounds used in this study.
interacting with cytosine–metalloporphyrin conjugates in a
membrane phase, we started by considering the association
constants reported by Ogoshi et al., corresponding to the
interaction of zinc octaethylporphyrin with 3Ј,5Ј-bis(O-triiso-
propylsilyl)deoxyribonucleosides in dichloromethane at 15 ЊC
(Table 1).20
Analysis of Ogoshi’s results reveals that the zinc()-contain-
ing metalloporphyrin ring has a greater preference for purine
nucleobases (i.e., adenine and guanine) than for pyrimidine-
type nucleobases (e.g., cytosine). Other studies also support
the notion that in the absence of a competing axial ligand,
metalloporphyrins containing a soft Lewis acidic metal center
are capable of interacting strongly with various amine-type
ligands,21–25 in particularly, nucleobases.20,26 To the extent that
such conclusions may be generalized, they lead to the consider-
ation that, in the absence of other effects, ISE containing metal-
loporphyrins would show selectivities dominated by specific
nucleobase metal coordination interactions. In the present
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1170

View Article Online
5Ј-GMP is additionally supported by results obtained using
the matched potential method. The matched potential method
was officially recommended by the IUPAC when the interfering
ions and/or the primary ion do not satisfy the Nernstian
condition.27 Using this latter method, the selectivity factor,
ksel5Ј-UMP/J (J represents interfering ions, in this case a competing
nucleotide) was found to follow the sequence:
2: 5Ј-AMP (Ϫ0.3) < 5Ј-GMP (0.1) < 5Ј-CMP (0.75)
3: 5Ј-AMP (Ϫ0.3) < 5Ј-GMP (Ϫ0.1) < 5Ј-CMP (0.2)
(using 5Ј-UMP at 10Ϫ4 M as the background, the concentration
of the interfering nucleotides was varied over the range
n × 10Ϫ3/n × 10Ϫ4 M).
The above findings are not consistent with what one would
infer from the Ogoshi study and led to the consideration that
factors other than nitrogen-centered metal–nucleotide axial
ligation or Watson–Crick base pairing (which would favor
5Ј-GMP) were acting to regulate the selectivity in the absence
of a lipophilic additive.
Among the interactions that could serve to regulate the
selectivity observed in the absence of a lipophilic additive, those
involving axial ligation were still thought to be the most
important. However, here it is important to appreciate that
binding modes other than direct nitrogen-centered co-
ordination could be important. For instance, binding of water
to the metalloporphyrin core (as an axial ligand), could reduce
or obviate the effect of direct metalloporphyrin–nucleobase
interaction. Such effects, which are not likely to be important
under the carefully controlled, anhydrous conditions associated
with the Ogoshi study, could be quite significant under the
interfacial ISE membrane conditions. Another effect, not
controlled in the Ogoshi study, is the metal–phosphate axial
ligation effect. The phosphate oxyanion is a potential hard
ligand and could compete with what are presumably softer
nucleobase nitrogen donors for the metal centers present in
2 and 3.
The importance of porphyrin metal core “deactivation” and,
indeed, more generally the importance of metalloporphyrin
axial ligand interactions, in terms of regulating binding affinity
and ISE membrane selectivity in systems such as 2 and 3,
was tested by exploring the preference, if any, the cytosine-
substituted Co()-porphyrin 2 displayed towards nucleobase
and phosphates. Using as analytes sodium dihydrophosphate
and diphenylphosphate, this receptor was found to display the
expected anionic responses, with sensitivities of Ϫ45 and Ϫ38
Fig. 2 Potentiometric response towards the nucleotides: 5Ј-AMP (∆),
5Ј-CMP (᭺), 5Ј-GMP (ᮀ), 5Ј-TMP (×) and 5Ј-UMP (ϩ) observed for
PVC-membranes based on 2 a) and 3 b) in the absence of a lipophilic
additive.
mV/decade with a linear range of 10Ϫ3–10Ϫ2 and 10Ϫ5–10Ϫ2
M
being recorded for the dihydrophosphate and diphenylphos-
phate salts, respectively. By contrast, when adenine phosphate
was chosen as the targeted analyte, a cationic response was
observed (sensitivity: ϩ38 mV/decade; linear range: 10Ϫ3–10Ϫ2
M). These findings lend important support for the notion that
this particular receptor can interact with a nucleotide via axial
ligation to both its constituent nucleobase and phosphate
subunits.
As a further means of assessing nitrogen vs. phosphate
ligation effects, and to establish a base-line for understanding
the role, if any, the appended cytosine hydrogen bond receptor
functionality present in 2 and 3 might play in regulating the
observed ISE selectivity, control experiments involving zinc()
tetraphenylporphyrin and cobalt() tetraphenylporphyrin were
carried out (Table 2).
The PVC-membranes derived from Co-tetraphenylporphyrin
showed the same ability to sense both nucleobase and various
phosphate species (dihydrophosphate, diphenylphosphate) as
did their cytosine substituted analogues. Likewise, the mem-
branes containing Zn-tetraphenylporphyrin demonstrated the
same weak response towards simple phosphorylated species
seen for the functionalized derivative, while demonstrating only
a cationic response towards nucleobases. In the case of the
instance, i.e., with receptors 2 and 3, such axial ligation effects
would be expected to be the most important selectivity deter-
mining factors in the absence of other competing binding
modes, or in the limit where the nucleobase-derived Watson–
Crick recognition effects are small. The question we sought to
address, therefore, was whether the latter would represent
important selectivity determinants and, if so, under what
conditions.
Under the conditions of our interfacial ISE experiments,
we found as a general rule that the PVC-membranes based on
the cytosine metalloporphyrin conjugates 2 and 3 displayed
a preference for pyrimidine over purine nucleotides in the
absence of a lipophilic additive (TDDMACl). In other words,
under these conditions, the pyrimidine nucleotides (5Ј-CMP;
5Ј-UMP, although not, however, 5Ј-TMP) engendered a greater
potentiometric response than the corresponding purine nucleo-
tides (5Ј-AMP; 5Ј-GMP). While it is difficult to discuss
the selectivity of ion-selective electrodes in the absence of a
Nernstian response towards the tested analytes, it is important
to note that our conclusion that the selectivity of cytosine-
substituted metalloporphyrins is larger for 5Ј-CMP than for
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1171

View Article Online
Table 2 Potentiometric responses of PVC-membranes based on cytosine-substituted Co()- and Zn()-porphyrins (2 and 3) and their unsubstituted
analogues (6 and 7) towards simple phosphate-type analytes
Analyte
2
6
3
7
Ϫ45 (10Ϫ3–10Ϫ2
Ϫ38 (10Ϫ5–10Ϫ2
ϩ38 (10Ϫ3–10Ϫ2
)
)
)
Ϫ36 (10Ϫ3–10Ϫ2
Ϫ42 (10Ϫ3–10Ϫ2
ϩ41 (10Ϫ3–10Ϫ2
)
)
)
Ϫ
a
a
a
a
a
H3PO4/H2PO4
Diphenylphosphate
Adenine phosphate
ϩ60 (10Ϫ3–10Ϫ2
)
a No detectable response.
Fig. 3 Schematic representation of the various proposed interactions: (a) expected ditopic binding of a targeted nucleotide, (b) axial ligation to the
metal (only), with no hydrogen bonds; (c) binding of a nucleotide substrate via only hydrogen bonds, and (d) “head-to-tail” dimer formation
involving only the receptor.
various nucleotides, the following selectivity order was observed
for the unsubstituted metalloporphyrins:
terms of determining both the extent of the ISE response and
the inherent nucleotide selectivities.
That a difference was seen between 2 and 3 and their
respective unfunctionalized controls provides support for the
notion that substitution with a nucleobase could serve to
impart a degree of nucleotide selectivity (Fig. 3a). Here, the
more hydrogen bonds the cytosine–substituent interaction is
capable of supporting, the greater the observed ISE signal
should be. In particular, if such effects are important, the
following selectivity order might be expected: 5Ј-AMP (× 1)
< 5Ј-TMP (× 2) ∼ 5Ј-UMP (× 2) ∼ 5Ј-CMP (× 2) ≤ 5Ј-GMP
(× 3) (the values in parentheses correspond to the number of
the possible hydrogen bonding interactions) (Fig. 3a, b). Of
course, such a prediction is predicated on the hydrogen bonding
interactions being the sole determinant of selectivity with axial
ligation, clearly an important factor, being completely ignored.
This is clearly an over-simplification hand, since a strong
predilection for nucleobase axial ligation is expected to reduce
the selectivity of a nucleobase-substituted metalloporphyrin
recognition element. Specifically, strong ligation effects are
expected to “wash out” the effects of Watson–Crick base pair-
ing (Fig. 3c).
6: 5Ј-TMP (0.00) < 5Ј-GMP (0.22) < 5Ј-CMP (0.25) < 5Ј-UMP
(0.38)
7: 5Ј-AMP (Ϫ0.31) < 5Ј-GMP (Ϫ0.13) < 5Ј-TMP (0.00) <
5Ј-CMP (0.08) < 5Ј-UMP (0.61),
where the values in parentheses refer to the selectivity factor
(ksel5Ј-TMP/J, J represents a competing nucleotide). As detailed in
the Experimental Section, because of the lack of Nernstian
response seen for the nucleotides in question, comparisons
of selectivity were made using the matched potential method.
Specifically, using 5Ј-TMP at 10Ϫ4 M as the background, the
concentration of the interfering anion was varied (up to
n × 10Ϫ3/n × 10Ϫ4 M) and the corresponding response recorded.
Thus, in contrast to their cytosine-substituted metallo-
porphyrin analogues, the simple unsubstituted systems display
a greater preference for 5Ј-UMP than for 5Ј-CMP. Further, for
the unsubstituted systems 6 and 7, the greatest distinction
between purine and pyrimidine analytes was observed in
the case of Zn()-tetraphenylporphyrin. While not fully
determined by the present experiments, presumably these
observations reflect the greater importance of nucleobase axial
ligation, as opposed to phosphate-based metal binding, in
Another complication that has to be considered is that self-
recognition, resulting in “head-to-tail” interactions between
separate molecules of 2 or 3 under the conditions of the
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1172

View Article Online
membrane-based experiments, could lead to the formation of
dimers or higher order aggregates (Fig. 3d). The formation of
such species, for which some support comes from mass
spectrometric analyses of 3 (cf. Experimental Section), would
necessarily reduce the ability of the nucleobase porphyrin
conjugates to function as nucleotide selective ISE sensing
elements.
Knowledge of the operative mechanism of the metallo-
porphyrin based sensors is critical since optimization of
membrane selectivity is highly dependent on the incorporation
of additional membrane components. Indeed, it has recently
been demonstrated that the presence of lipophilic electrically
charged additives improves the potentiometric behavior of
certain anion-selective electrodes, including ones prepared with
metalloporphyrins and cobyrinates.28–30 The type of additive
required depends on the carrier mechanism involved. It is
known that the oxidation state of the metal center within the
metalloporphyrin structure can dictate the possible mechanism
for the interaction of anions with such a receptor at the sample/
membrane interface.31 Porphyrins with metal() centers can
function via a neutral carrier mechanism,32 as such, their
response sensitivity could be counter cation limited. For this
reason, neutral metalloporphyrin-based membranes containing
lipophilic cationic additives should exhibit improved selectiv-
ity.30 In the present instance, involving the cytosine substituted
porphyrins, 2 and 3, the use of a lipophilic additive is expected
to be beneficial since it would reduce the polarity of the
environment and facilitate charge compensation. In doing so, it
should enhance the importance of the Watson–Crick base pair-
ing interactions, which should lead to selectivity for 5Ј-GMP in
those cases where these latter are sufficiently large.
As a test of the above expectations, we investigated the
effect of adding a lipophilic positive charged additive (i.e.,
TDDMACl) into the ISE membranes containing receptor 2.
Furthermore, to test whether this latter system was indeed
acting as an efficient ditopic receptor for 5Ј-GMP, we decided to
use this latter nucleotide as the targeted substrate.
The optimal molar ration of the ionic additive was found to
be 10 mol% with respect to the total ionophore concentration
present in the membrane (Fig. 4). As compared to the
membranes made up without any cationic additive, those
membranes containing 10 mol% TDDMACl demonstrated
considerably higher sensitivity (in particular, Ϫ26 and Ϫ29
mV/decade in the concentration range 10Ϫ3–10Ϫ2 M for 5Ј-AMP
and 5Ј-GMP, respectively) and selectivity for 5Ј-GMP over
Fig. 4 Influence of lipophilic additive content on (a) sensitivity and
(b) potentiometric selectivity coefficients (KPot.5Ј-GMP/5Ј-AMP) of a PVC-
membrane based on cytosine-substituted Co()-porphyrin conjugate 2
toward 5Ј-GMP.
underscores the utility of a basic, molecular recognition
inspired approach to sensor design but also serves to highlight
the difficulties associated with extrapolating simple well-defined
solution phase-based results into the more “real world” con-
ditions associated with ISE sensor development.
5Ј-AMP (KPot.
= 0.045). It is interesting to note that
5Ј-GMP/5Ј-AMP
the resulting value KPot.
is approximately one order
5Ј-GMP/5Ј-AMP
greater than that reported by Umezawa et al. (KPot.
=
5Ј-GMP/5Ј-AMP
0.45).9 An increase in the TDDMACl content from 50 mol%
to 150 mol% led to a significant decrease in selectivity; under
these latter conditions, the membrane began to function as an
ordinary ion-exchanger (Fig. 4b). However, this same drop off
in selectivity serves to highlight the fact that under appro-
priately chosen conditions (e.g., 10 mol% TDDMACl), the
cytosine-substituted metalloporphyrin systems described in this
report, in particular, the Co()-containing complex 2, can be
made to function as signaling agents for their Watson–Crick
complement, 5Ј-GMP.
Experimental
Materials
All reagents were of the highest grade commercially available
and used without further purification. Poly(vinyl chloride) high
molecular weight (PVC), 2-nitrophenyl octyl ether (o-NPOE),
tridodecylmethylammonium chloride (TDDMACl), tetra-
hydrofuran (THF; stored over 3 Å molecular sieves) were
purchased from Fluka (Switzerland). Adenosine 5Ј-mono-
phosphate sodium salt (5Ј-AMP), cytidine 5Ј-monophosphate
disodium salt (5Ј-CMP), guanosine 5Ј-monophosphate di-
sodium salt (5Ј-GMP), thymidine 5Ј-monophosphate sodium
salt (5Ј-TMP), uridine 5Ј-monophosphate disodium salt (5Ј-
UMP), 2-[4-(2-hydroxyethyl)-1-piperazine]ethanesulfonic acid
(HEPES), cobalt and zinc tetraphenylporphyrins were from
Sigma-Aldrich Chemie (Steinheim, Germany). Adenine phos-
phate salt and diphenylphosphate were obtained from Sigma
and Aldrich Chem. Co. (Germany), respectively. Distilled water
was used to prepare buffer and standard solutions.
Conclusion
In summary, we have described a new class of ditopic receptors,
based on the covalent linking of a nucleobase, cytosine in the
present instance, to a metalloporphyrin. While a number of
competing interactions serve to complicate the use of these
systems as sensing elements in ISEs, under appropriately
chosen conditions (10 mol% TDDMACl as a lipophilic
additive; use of cobalt() as the coordinated porphyrin metal
center), good selectivities for the Watson–Crick complementary
analyte, 5Ј-GMP, can be obtained. The present work thus
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1173

View Article Online
Electrode preparation and ISE measurements
followed by removal of the solvent yielded the protected version
of conjugate 1 in 89% yield. This intermediate product was not
isolated or characterized but rather was immediately subjected
to deprotection using the procedure described earlier;8 product
1 was obtained by crystallization from MeOH–DCM diethyl
ether in 92% yield. This cytosine–porphyrin conjugate was used
as a precursor for the corresponding Co() and Zn() metal
complexes, 2 and 3; these latter were prepared in quantitative
yield by heating the metal free form, 1, with the appropriate
acetoacetonates in chloroform–methanol 1 : 1 at reflux for
4 hours.
Ion-selective membranes were prepared in accord with the
procedure described in ref. 12. In the present study, 0.7 ml THF
was used to dissolve approximately 100 mg of a mixture con-
taining 3 wt% of the receptor in question, 22 wt% PVC, and 75
wt% o-NPOE. The resulting membranes, obtained following
evaporation as before, were mounted on an electrode body
(Crytur, Czech Republic). The influence of cationic sites on the
potentiometric characteristics of these membranes was assessed
by studying not only these membranes but ones containing,
in addition, 10, 50 and 150 mol% TDDMACl relative to the
receptor (incorporated at 3 wt% level, as noted above). Control
electrodes, containing just Co-tetraphenylporphyrin 6 and
Zn-tetraphenylporphyrin 7, were prepared and studied as
reported previously.12 EMF measurements were performed
using a digital voltammeter, Model M1T330 (Metra s.p.,
Blansko, Czech Republic), and in accord with the following cell
assembly: Hg | Hg2Cl2 | 3 M KCl || 0.1 M HEPES–NaOH pH
6.6 || sample | modified PVC-membrane | 0.1 M KCl | AgCl | Ag.
All potentiometric analyses were carried out at ambient tem-
perature. The pH was monitored using a glass electrode Type
01-29 B (Labio Prague, Czech Republic) and a Type OP-205/1
pH-meter (Radekis, Budapest, Hungary). In the studies of
potentiometric response and anion selectivity, working solu-
tions of the analytes in question were prepared by diluting con-
centrated stock solutions with 0.1 M HEPES adjusted to pH
6.6 with NaOH. Calibration curves were constructed by plot-
ting the potential vs. the logarithm of the concentration of the
anion present in the buffer solution. Anion concentrations
rather than activities were used because it is difficult to estimate
activity coefficients in the zwitterionic buffer. Before starting
the ISE studies, the electrodes were soaked overnight in HEPES
buffer (0.1 M HEPES, adjusted to pH 6.6 by the addition of
NaOH as above) in the presence of the analyte. Potentiometric
selectivity coefficients (log KPot.I/J) were then determined by the
separate solution method,27 with the primary (I) and interfering
(J) ion concentrations being 1.0 × 10Ϫ2 M for both the PVC
membranes containing the receptors (e.g., 2) and the analyte
(e.g., 5Ј-GMP). In certain instances, especially those wherein a
non-Nernstian response was observed, the selectivity factor
(kselI/J) was measured using the matched potential method.27
Cytosine–porphyrin conjugate 1. 1H NMR (500 MHz, CDCl3)
δ 10.21 (s, 4H, CH), 10.14 (s, 4H, CH), 10.13 (s, 4H, CH), 10.11
(s, 4H, CH), 6.72 (d, 1H, cytosine C6H), 4.75 (d, 1H, cytosine
C5H), 4.44 (br t, 2H, CH2CH2CON), 4.1 (t, 8H, 4 × CH3CH2),
4.75 (br s, 2H, N–CH2-bis(t-Bu)phenyl), 3.669, 3.665, 3.605
(3 × s, 9H, 3 × CH3), 3.38 (br, 2H, CO–N–CH2CH2–N–cyt),
3.249 (br t, 2H, CH2CH2CON), 3.139 (br, 2H, CO–N–
CH2CH2–N–cyt), 1.931, 1.89 (2 × br q, 12H, 4 × CH3CH2), 1.25
(s, 18H, t-Bu), Ϫ3.792 (s, 2H, 2 × NH). 13C NMR (125 MHz,
CDCl3 with 5% CD3OD) δ 12.30, 18.42, 19.22, 20.59, 22.95,
30.42, 32.10, 35.86, 46.65, 48.56, 54.09, 76.37, 91.96, 97.28,
120.54, 122.20, 136.20, 144.98, 152.17, 163.40, 173.74. HR FAB
MS: calcd. for C55H71N8O2: 875.569999; found 875.569265.
Elemental analysis: calcd. for C55H70N8O2 (875.20): C 75.48, H
8.06, N 12.80; found C 75.23, H 8.18, N 12.60%.
Receptor 2. HR FAB MS: calcd. for C55H68N8O2Co:
931.479722; found: 931.478546. Elemental analysis: calcd. for
C55H68N8O2 (932.11): C 70.87, H 7.35, N 12.02; found C 70.63,
H 7.12, N 11.89%.
Receptor 3. HR FAB MS: calcd. for C55H68N8O2Zn:
936.475669; found: 936.476541. Elemental analysis: calcd. for
C55H68N8O2Zn (938.57): C 70.38, H 7.30, N 11.94; found C
70.10, H 7.12, N 11.69%. Dimer of 3, as inferred from the HR
FAB mass spectrum: Calcd for C110H137N16O4Zn2: 1873.969164;
found: 1873.956841.
Acknowledgements
This work was supported by the Ministry of Education, Youth
and Sports of the Czech Republic (grant no. CEZ: MSM
223400008), the Grant Agency of the Czech Republic (grant
nos. 203/02/0933, 203/02/0420), the EU (grant QLRT-2000-
02360), and the National Institutes of Health (grant no. GM
58907).
Instrumentation for receptor characterization
NMR spectra were obtained using a General Electric QE-300
spectrometer at the University of Texas at Austin. High
resolution fast atom bombardment mass spectra (HR FAB MS)
were recorded at the University of Texas at Austin Department
of Chemistry and Biochemistry MS Facility using a VG
ZAB-2E instrument.
References
Synthesis and characterization
1 Y. Umezawa, M. Kataoka, W. Takami, E. Kimura, T. Koike and
H. Nada, Anal. Chem., 1988, 60, 2392–2396.
General procedure for preparing cytosine–metalloporphyrin
conjugates, 1–3. Compounds 1–3 were prepared via amide bond
formation starting from the porphyrin carboxylic acid 4, pre-
pared according to the procedure of ref. 33, and the protected
aminoethylcytosine 5.13,34 Thus, 4 (210 mg; 3.98 × 10Ϫ4 mol) was
dissolved in 5 ml of dry dimethylformamide (DMF) and cooled
to 0 ЊC. To this solution was then added 0.75 ml diisopropyl-
carbodiimide. Stirring was continued at 0 ЊC for one hour,
before the mixture was allowed to warm to RT. At this point,
5 (465 mg; 7.78 × 10Ϫ4 mol) was added along with 5 mg
1-hydroxybenzotriazole (HOBT) and 30 mg 4-dimethyl-
aminopyridine (DMAP). The reaction mixture was then stirred
in the dark for 30 hours under argon, before the DMF was
removed in vacuo. The resulting residue was dissolved in
dichloromethane (DCM) and subjected to column chromato-
graphy over silica gel, using DCM containing 0–5% methanol
(MeOH) as the eluent. Isolation of the appropriate fraction,
2 H. Furuta, K. Furuta and J. L. Sessler, J. Am. Chem. Soc., 1991, 113,
4706–4707.
3 K. Tohda, M. Tange, K. Odashima, Y. Umezawa, H. Furuta and
J. L. Sessler, Anal. Chem., 1992, 64, 960–964.
4 V. Král, J. L. Sessler and H. Furuta, J. Am. Chem. Soc., 1992, 114,
8704–8705.
5 J. L. Sessler, H. Furuta and V. Král, Supramol. Chem., 1993, 1,
209–220.
6 K. Tohda, R. Naganawa, X. M. Lin, M. Tange, K. Umezawa,
K. Odashima, Y. Umezawa, H. Furuta and J. L. Sessler, Sens.
Actuators, B, 1993, 13–14, 669–672.
7 K. Odashima, R. Naganawa, H. Radecka, M. Kataoka, E. Kimura,
T. Koike, K. Tohda, M. Tange, H. Furuta, J. L. Sessler, K. Yagi and
Y. Umezawa, Supramol. Chem., 1994, 4, 101–113.
8 V. Král and J. L. Sessler, Tetrahedron, 1995, 51, 539–554.
9 S. Amemiya, P. Buhlmann, K. Tohda and Y. Umezawa, Anal. Chim.
Acta, 1997, 341, 129–139.
10 S. Aoki, Y. Honda and E. Kimura, J. Am. Chem. Soc., 1998, 120,
10018–10026.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1174

View Article Online
11 P. Buhlmann, S. Amemiya, S. Nishizawa, K. P. Xiao and
Y. Umezawa, J. Inclusion Phenom. Mol. Recognit. Chem., 1998, 32,
151–163.
12 V. Král, J. L. Sessler, T. V. Shishkanova, P. A. Gale and R. Volf,
J. Am. Chem. Soc., 1999, 121, 8771–8775.
23 T. Mizutani, K. Wada and S. Kitagawa, J. Org. Chem., 2000, 65,
6097–6106.
24 J. A. J. Brunink, C. Di Natale, F. Bungaro, F. A. M. Davide,
A. D’Amico, R. Paolesse, T. Boschi, M. Faccio and G. Ferri, Anal.
Chim. Acta, 1996, 325, 53–64.
13 J. L. Sessler, V. Král, T. V. Shishkanova and P. A. Gale, Proc. Natl.
Acad. Sci. USA, 2002, 99, 4848–4853.
14 S. Picioreanu, I. Poels, J. Frank, J. C. van Dam, G. W. K. van Dedem
and L. J. Nagels, Anal. Chem., 2000, 72, 2029–2034.
15 M. Sirish and H.-J. Schneider, J. Am. Chem. Soc., 2000, 122,
5881–5882.
25 D. Delmarre and C. Bied-Charreton, Sens. Actuators, B, 2000, 62,
136–142.
26 Y. Kuroda, H. Hatakeyama, H. Seshimo and H. Ogoshi, Supramol.
Chem., 1994, 3, 267–271.
27 Y. Umezawa, K. Umezawa and H. Sato, Pure Appl. Chem., 1995, 67,
507–518.
16 S. Aoki and E. Kimura, J. Am. Chem. Soc., 2000, 122,
28 M. Rothmaier and W. Simon, Anal. Chim. Acta, 1993, 271, 135–141.
29 U. Schaller, E. Bakker, U. E. Spichiger and E. Pretsch, Anal. Chem.,
1994, 66, 391–398.
30 E. Bakker, E. Malinowska, R. D. Schiller and M. E. Meyerhoff,
Talanta, 1994, 41, 881–890.
31 H. Yim, C. E. Kibbey, S. Ma, D. M. Kliza, D. Lu, S. B. Park,
C. Espadas-Torre and M. E. Meyerhoff, Biosens. Bioelectron., 1993,
8, 1–38.
4542–4548.
17 C. Anda, A. Llobet, V. Salvado, J. Reibenspies, R. J. Motekaitis and
A. E. Martell, Inorg. Chem., 2000, 39, 2986–2999.
18 S. Lubomir, F. Diederich and V. Gramlich, Helv. Chim. Acta, 2000,
83, 93–113.
19 J. L. Sessler, J. W. Genge, V. Král and B. L. Iverson, Supramol.
Chem., 1996, 8, 45–52.
20 H. Ogoshi, H. Hatakeyama, J. Kotani, A. Kawashima and
Y. Kuroda, J. Am. Chem. Soc., 1991, 113, 8181–8183.
21 D. W. J. Mccallien, P. L. Burn and H. L. Anderson, J. Chem. Soc.,
Perkin Trans. 1, 1997, 17, 2581–2586.
22 M. Inamo and I. Yoneda, Inorg. Chem. Commun., 1999, 2,
331–333.
32 M. Huser, W. E. Morf, K. Fluri, K. Seiler, P. Schulthess and
W. Simon, Helv. Chim. Acta, 1990, 73, 1481–1496.
33 J. L. Sessler, J. W. Genge, A. Urbach and P. Sansom, Synlett, 1996, 2,
187–188.
34 H. Furuta, D. Magda and J. L. Sessler, J. Am. Chem. Soc., 1991, 113,
978–985.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 6 9 – 1 1 7 5
1175