Synthetic model of the phosphate binding protein: Solid-state structure and solution-phase anion binding properties of a large oligopyrrolic macrocycle

Article abstract of DOI:10.1021/jo071106g

The macrocyclic receptors 4-6 were synthesized via the anion-templated condensation of appropriately chosen dialdehyde and diamine building blocks. Whereas all three products could be obtained directly via the appropriate choice of reaction conditions, th

Full text of DOI:10.1021/jo071106g

Synthetic Model of the Phosphate Binding Protein: Solid-State
Structure and Solution-Phase Anion Binding Properties of a Large
Oligopyrrolic Macrocycle
Evgeny A. Katayev,*,† Jonathan L. Sessler,*,‡ Victor N. Khrustalev,† and Yuri A. Ustynyuk§
A. N. NesmeyanoV Institute of Organoelement Compounds, Russian Academy of Sciences, 28, VaViloV
Street, 119991, Moscow, Russian Federation, Department of Chemistry and Biochemistry and Institute for
Cellular and Molecular Biology, 1 UniVersity Station-A5300, UniVersity of Texas at Austin, Austin, Texas
78712-0165, and Department of Chemistry, M. V. LomonosoV Moscow State UniVersity, Leninskie Gory,
119992 Moscow, Russian Federation
katayeV@ineos.ac.ru; sessler@mail.utexas.edu
ReceiVed May 31, 2007
The macrocyclic receptors 4-6 were synthesized via the anion-templated condensation of appropriately
chosen dialdehyde and diamine building blocks. Whereas all three products could be obtained directly
via the appropriate choice of reaction conditions, the larger [3+3] product, 6, which incorporates three
of each precursor subunit, could also be obtained conveniently via an indirect procedure involving ring
expansion of the smaller [2+2] macrocycle 4. As detailed earlier (Sessler, J. L.; Katayev, E. A.; Pantos,
G. D.; Reshetova, M. D.; Khrustalev, V. N.; Lynch, V. M.; Ustynyuk, Y. A. Angew. Chem. 2005, 117,
7552-7556; Angew. Chem., Int. Ed. 2005, 44, 7386-7390), this ring expansion occurs under
thermodynamic control in the presence of HSO4- and H2PO4- anions in acetonitrile solution and serves
2-
to effect the conversion of 4 to 6. An analysis of the X-ray crystal structure of complex 6H22+‚HPO4
revealed a strong resemblance to the active site of the phosphate binding protein (PBP) with similar
structural analogies being drawn between the active site of the sulfate binding protein (SBP) and the
corresponding hydrogensulfate anion complex. In both cases, the anions are bound in a 1:1 fashion in
the solid state through a complementary hydrogen bond network involving both the receptor 6 and the
anions. UV-vis spectroscopic titrations provide support for the conclusion that macrocycle 6 binds the
hydrogensulfate and dihydrogenphosphate anion (studied as the corresponding tetrabutylammonium salts)
with high selectivity and affinity in acetonitrile (log Ka for the first binding interaction approaching 7),
-
-
albeit with different receptor-to-anion binding stoichiometries (1:1 vs 1:3 for HSO4 and H2PO4 ,
respectively).
Introduction
sensors.1 The anion-host interaction principles elucidated in
the course of these studies have served to animate advances in
such disparate areas as sensor technology,2 waste remediation,3
catalysis, and drug discovery.4,5 Meanwhile, a specific objective
Ever since the anion recognition properties of simple organic
molecules were noted more than 30 years ago, considerable
effort has been devoted to the synthesis of anion receptors and
† Russian Academy of Science.
(1) Sessler, J. L.; Gale, P. A.; Cho, W.-S. Anion Receptor Chemistry;
Royal Society of Chemistry: Letchworth, UK, 2006.
(2) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed. 2001, 40, 486-516.
‡ University of Texas at Austin.
§ M. V. Lomonosov Moscow State University.
10.1021/jo071106g CCC: $37.00 © 2007 American Chemical Society
Published on Web 08/18/2007
7244
J. Org. Chem. 2007, 72, 7244-7252

Synthetic Model of the Phosphate Binding Protein
has been to develop receptors that can mimic the anion
recognition motifs found in natural systems,6 including the
phosphate- and sulfate-binding proteins (PBP and SBP), protein
kinases, and phosphatases. This is because an increased
understanding of the underlying anion recognition features
exploited by Nature is expected to lead to improvements in
structure-based drug design, as well as better modeling of the
natural enzymatic processes.7 Recently we reported the synthesis
of receptors induced by using the conjugate acid of a specifically
targeted anion as a template for macrocycle formation. Two
phosphate selective receptors 18 and 49 were prepared in this
way and were shown to possess binding affinities of up to log
Ka ) 5 in acetonitrile for the first binding event with an overall
1:2 (receptor to dihydrogenphosphate anion) binding stoichi-
ometry. In this paper we report a template-mediated synthesis
of the larger system 6 and a detailed analysis of the hydrogen-
bonding networks present in its phosphoric acid and sulfuric
acid salts. We also report the first synthesis of a reduced
analogue of 4 (macrocycle 5), efforts to produce the smaller
congeners 2 and 3, as well as solution phase anion binding
studies involving all three larger bipyrrole-containing receptors
(4-6). On the basis of the structural similarity (cf. Table 1)
between the phosphoric acid salt 6‚H3PO4, which is seen to be
incorporate both H-donor and acceptor sites within the same
overall receptor framework. This strategy, which provides a
complement to the recent work of Bowman-James18 and Gale,19
was inspired by the specific challenge of creating a synthetic
model for the active center of PBP. However, in contrast to the
approach adopted by others, in our case imine bonds, rather
than the amide functionality, were used to link the various
“building blocks”. This was done so as to build on our previous
efforts in the expanded porphyrin area and to produce systems
that are potentially amenable to acid-catalyzed, template-
mediated formation. This choice was also expected to produce
macrocyclic products that contain ancillary H-bond acceptor
sites. Since inorganic phosphate and its derivatives can exist in
several protonated forms (e.g., H2PO4- and HPO42-) at or near
physiological pH, it was specifically thought that the use of
imine linkers, which might be subject to protonation, would
allow for the accommodation of these different forms.
In this work, 3,3′-dimethyl-4,4′-dipropyl-5,5′-diformyl-2,2′-
bipyrrole 7 and the diamine building blocks 8 and 9 were used
to produce the small combinatorial library of receptors 2-6
(Figure 1). The selection of ostensibly similar amide (8)- and
amine (9)-linked o-phenylenediamine-derived precursors was
made so as to be able to assess the importance of the putative
H-bond acceptor binding sites in the resulting receptors (i.e.,
through use of the aminomethyl precursor, 9, and amide
precursor, 8, respectively).20 Since these precursors also differ
in their inherent rigidity, we were likewise curious to see if
their use would give rise to different ratios of combinatorial
products (e.g., [1+1], [2+2], [3+3], etc.) incorporating different
numbers of precursor subunits, under otherwise identical ring-
forming conditions.
2-
the dihydrogenphosphate complex, 6H22+‚HPO4 in the solid
state, and the binding site seen in the PBP,5, 10, 11 as well as the
recognized ability of its free base form to bind phosphate anions
in acetonitrile solution, we conclude that receptor 6 may be
considered as a rudimentary synthetic model for this naturally
occurring anion binding protein.
Results and Discussion
As in the case of our previous studies, acid-promoted imine
formation was chosen to effect macrocyclization. Under the
conditions employed (precursor concentrations of ca. 0.1-0.15
M; 2 molar equiv of acid; methanol as solvent), this reaction is
inherently reversible, and was thus expected to depend not only
on the choice of precursor but also on the choice of acid catalyst.
In accord with such expectations, initial screening studies
revealed that only phosphoric and sulfuric acids gave rise to
the larger macrocyclic products 4-6 in good yields. The use
of other acids, including HCl, HBr, CF3COOH, HNO3, HClO4,
and HReO4, gave rise to mixtures of products 2-6 as deduced
from MALDI-TOF or ESI mass spectrometric analysis. On the
basis of these findings, we infer that phosphoric and sulfuric
acid work well as templates for the clean formation of 4-6
because the corresponding conjugate anions fit well within the
hydrogen-bonding network provided by the incipient macrocy-
clic receptor. Within the context of this overall conclusion, it
is of interest to note that the use of sulfuric acid in the reaction
of building blocks 7 and 8 leads to the formation of the bis-
sulfuric acid salt of 4 (the [2+2] condensation product) in almost
90% yield, whereas the use of phosphoric acid leads to the
formation of the monophosphoric salt of macrocycle 6 (the
[3+3] product) in 46% yield (Scheme 1). Interestingly, the
[1+1] product, compound 2 (or its protonated derivatives), could
General aspects of tetrahedral oxoanion receptors have
recently been reviewed by us6 and others.12 In the case of
phosphate anions, charged polyamine13 and guanidine14 recep-
tors, protonated expanded porphyrins,15 and various zinc- and
copper-based complexes16 have proved to be the most efficient
systems. In recent studies we have found that appropriately
designed pyrrole-pyridine hybrid macrocycles show promise
as phosphate anion receptors.8,17 These systems, built up from
pyrrolic and amidopyridine precursors, were designed to
(3) Bianchi, A.; Bowman-James, K.; Garcia-Espan˜a, E. Supramolecular
Chemistry of Anions; Wiley-VCH, New York 1997.
(4) Darbre, T.; Reymond, J.-L. Acc. Chem. Res. 2006, 39, 925-934.
(5) Morales, R.; Berna, A.; Carpentier, P.; Contreras-Martel, C.; Renault,
F.; Nicodeme, M.; Chesne-Seck, M. L.; Bernier, F.; Dupuy, J.; Schaeffer,
C.; Diemer, H.; Van-Dorsselaer, A.; Fontecilla-Camps, J. C.; Masson, P.;
Rochu, D.; Chabriere, E. Structure 2006, 14, 601-609.
(6) Katayev, E. A.; Ustynyuk, Y. A.; Sessler, J. L. Coord. Chem. ReV.
2006, 250, 3004-3037.
(7) Hirsch, A. K. H.; Fischer, F. R.; Diederich, F. Angew. Chem., Int.
Ed. 2007, 46, 338-352.
(8) Sessler, J. L.; Katayev, E.; Pantos, G. D.; Ustynyuk, Y. A. Chem.
Commun. 2004, 1276-1277.
(9) Katayev, E. A.; Pantos, G. D.; Reshetova, M. D.; Khrustalev, V. N.;
Lynch, V. M.; Ustynyuk, Y. A.; Sessler, J. L. Angew. Chem., Int. Ed. 2005,
44, 7386-7390.
(10) Luecke, H.; Quiocho, F. A. Nature 1990, 347, 402-406.
(11) Yao, N.; Ledvina, P. S.; Choudhary, A.; Quiocho, F. A. Biochemisrty
1996, 35, 2079-2085.
(17) Katayev, E. A.; Boev, N. V.; Khrustalev, V. N.; Ustynyuk, Y. A.;
Tananaev, I. G.; Sessler, J. L. J. Org. Chem. 2007, 72, 2886-2896.
(18) Kang, S. O.; Begum, R. A.; Bowman-James, K. Angew. Chem.,
Int. Ed. 2006, 45, 7882-7894.
(19) Brooks, S. J.; Gale, P. A.; Light, M. E. Supramol. Chem. 2007, 19,
9-15.
(20) Sessler, J. L.; Katayev, E.; Pantos, G. D.; Scherbakov, P.; Reshetova,
M. D.; Khrustalev, V. N.; Lynch, V. M.; Ustynyuk, Y. A. J. Am. Chem.
Soc. 2005, 127, 11442-11446.
(12) Kang, S. O.; Hossain, M. A.; Bowman-James, K. Coord. Chem.
ReV. 2006, 250, 3038-3052.
(13) Llinares, J. M.; Powell, D.; Bowman-James, K. Coord. Chem. ReV.
2003, 240, 57-75.
(14) Schmidtchen, F. P. Coord. Chem. ReV. 2006, 250, 2916-2928.
(15) Sessler, J. L.; Davis, J. M. Acc. Chem. Res. 2001, 34, 989-997.
(16) O’Neil, E. J.; Smith, B. D. Coord. Chem. ReV. 2006, 250, 3068-
3080.
J. Org. Chem, Vol. 72, No. 19, 2007 7245

Katayev et al.
FIGURE 1. Hybrid macrocycles germane to the present study.
SCHEME 1. Products of Sulfuric and Phosphoric Acid Templated Cyclizations of Building Blocks 7-9
not be obtained in isolable quantities by using either procedure.
When the pyridine-2,6-diaminomethyl containing building block
9 was used, the use of phosphoric acid produced the [2+2]
product 5 in 72% yield, while sulfuric acid gave rise to a mixture
of the acid salts of [1+1] (3) and [2+2] (5) condensation
products (Scheme 1) that did not prove amenable to ready
separation. Under neither set of conditions were appreciable
quantities of [3+3] product, corresponding to a pyridine-
containing analogue of 6 (structure not shown), observed. On
this basis it is concluded that both the choice of acid template
and the nature of the pyridine-containing precursors are critical
to defining the nature of the products produced under the present
imine-forming macrocyclization conditions.
Once produced, the free receptors could be obtained via
neutralization of the acid salts with an excess of triethylamine
in methanol solution. In general, these free-base forms are stable
in solution for 2-3 days under normal laboratory conditions or
for months in the solid phase. However, when the free receptor
form of the [2+2] product 4 was allowed to stand for 5 days in
acetonitrile in the presence of hydrogensulfate or dihydrogen-
phosphate anion (in the form of the corresponding tetrabuty-
lammonium salts), followed by crystallization, the larger [3+3]
product 6 was isolated.9 Although cation-mediated expansions
of imine-derived macrocyclic receptors are known, having been
studied in some detail recently by Love and co-workers,21
corresponding anion-mediated ring expansions have only re-
cently been shown to be possible.9
No expansion of macrocycle 4 was observed when an excess
of ammonium sulfate was used in place of tetrabutylammonium
hydrogensulfate, again in acetonitrile solution. On this basis,
we conclude that the presence of protons (from, e.g., HSO4- or
H2PO4-) is required to induce what must be a thermodynami-
cally driven reorganization of the receptor via solvolysis of the
imino bonds. To the extent such a conclusion is correct, it
implies that, at least in acetonitrile, receptor 6 has a higher
(21) Givaja, G.; Blake, A. J.; Wilson, C.; Schroder, M.; Love, J. B. Chem.
Commun. 2005, 4423-4425.
7246 J. Org. Chem., Vol. 72, No. 19, 2007

Synthetic Model of the Phosphate Binding Protein
FIGURE 2. ORTEP POV-ray rendering of the structure of 6‚H3PO4 as deduced from a single-crystal X-ray diffraction analysis: (a) view along
the c-axis and (b) view along the a-axis. Ellipsoids are shown at the 20% probability level.
affinity for the hydrogensulfate and dihydrogenphosphate anions
than does its smaller congener 4. One of the objectives of the
present study was to test the extent to which such a conclusion
is correct.
receptor 6 is doubly twisted in the case of this structure, just as
it was in the previous one involving 6H22+‚SO42-. The P-O
distances are different and equal to 1.497(4), 1.500(4), 1.503-
(4), and 1.625(5) Å (cf. the Experimental Section in the
Supporting Information). These distances compare to the
corresponding S-O parameters of 1.470(4), 1.475(4), 1.458-
Solid-State Structural Analyses and Comparisons with the
Active Sites of the PBP and SPB. Prior to carrying out
solution-phase studies, we sought to obtain insights from solid-
state structural analyses. Although we succeeded in obtaining
diffraction grade crystals for 4 and 6, unfortunately this success
did not translate into the other members of the present series.
Macrocycles such as 1-6 have proved difficult to crystallize,
in part because the acid salts display poor solubility in all
common organic solvents and water. Macrocycle 4 was thus
crystallized as solvate with 2 molecules of water encapsulated
within the figure-8 shaped core, while macrocycle 6 was first
crystallized as its sulfuric acid complex, 6‚H2SO4, as was
discussed by us recently,9 and now as the phosphoric acid salt
in the context of this work. In both structures involving 6, the
solid-state species are best considered as complexes of the
corresponding dianion in the solid state (i.e., 6H22+‚SO42- and
6H22+‚HPO42-, respectively), as inferred from the overall
structure and from the fact that the hydrogen atoms bound to
nitrogen and phosphorus atoms were specifically observed in
the ∆F map for the new structure of 6‚H3PO4. For both
structures, which are nearly identical, there are two crystallo-
graphically independent molecules in the unit cell, which reflect
different conformers. From an inspection of Figure 2, which
shows the structure of 6H22+‚HPO42-, one can see that the large
macrocycle is twisted around the anion, forming a tight hydrogen
bond network. Three bipyrrole subunits surround the anion. Two
of these subunits act as cross sections for the macrocycle, while
the other defines the bottom of the receptor (Figure 2b). To the
right and left sides of the anion, there are two diamidopyridine
subunits that are oriented parallel to the bipyrrole units. From
an analysis of the X-ray structural parameters, it is inferred that
2-
(4), and 1.459(4) Å seen in the structure of 6H22+‚SO4
reported earlier.9
To facilitate a comparison of the hydrogen bond networks
that stabilize1 the bound anions in the solid structures of 6‚H3-
PO4 and 6‚H2SO4 with those present at the active sites of the
SBP22 and PBP,10 a schematic representation of the respective
anion coordination motifs is given in Figure 3. Relevant
structural parameters for the hydrogen bonding interactions are
given for all four systems in Tables 1 and 2. As can be seen
from this drawing and inferred from these tables, in all four
structures the average distances between the atoms involved in
the hydrogen bonding interactions are very similar to one another
and can be considered the same within the error limits. The
average angles of the individual hydrogen bonds are slightly
smaller for the model systems than for the enzymes, a disparity
that is ascribed to the rigidity of the macrocycle. In the synthetic
systems a network of hydrogen bond donors and acceptors is
established via a combination of three 2-iminobipyrrole subunits,
two 2,6-diamidopyridine fragments, and two protonated imine
residues per bound anion. In the case of the hydrogenphosphate
complex, 6H22+‚HPO42- (6‚H3PO4), one imine nitrogen acts as
a hydrogen bond acceptor and interacts with O(4)H from the
anion. In contrast, the other imine interacts with O(3). Thus,
the bound hydrogenphosphate anion is stabilized by one more
hydrogen bonding interaction than the sulfate anion present in
the structure of 6‚H2SO4 (12 vs 11). The principle difference
between these two structures thus lies in O4 coordination mode
(22) Pffugrath, J. W.; Quiocho, F. A. Nature 1985, 314, 257-260.
J. Org. Chem, Vol. 72, No. 19, 2007 7247

Katayev et al.
FIGURE 3. Schematic representations of the hydrogen bond networks present in complexes 6H22+‚HPO42- and 6H22+‚SO42- (structures 6‚H3PO4
and 6‚H2SO4, respectively) and the active sites of the PBP and SBP.
TABLE 1. Geometric Parameters Corresponding to the Anion-Stabilizing Hydrogen Bond Interactions in the PBP10 and 6‚H3PO4
PBP
6‚H3PO4
distance (Å)
phosphate
atom
distance (Å)
X‚‚‚Y
angle (deg)
X‚‚‚H‚‚‚Y
phosphate
atom
angle (deg)
X‚‚‚H‚‚‚Y
Y (atom)
Y (atom)
X‚‚‚Y
O1
NH(Thr10)
OH(Thr10)
NH(Arg135)
NH(Arg135)
OH(Ser139)
NH(Thr141)
OH(Thr141)
NH(Ser38)
OH(Ser38)
NH(Gly140)
NH(Phe11)
O(Asp56)
2.81
2.69
2.84
2.90
2.77
2.83
2.68
2.67
2.63
2.76
2.92
2.45
156.0
167.2
172.3
148.9
169.0
159.2
149.7
163.9
160.9
177.1
163.3
151.2
O1
N1
N2
N6
N15
N16
N17
2.825
2.812
3.011
2.947
2.714
2.936
156
148
107
160
162
163
O2
O2
O3
O4
O3
O4
N7
N8
N9
N20
N18
N21
2.799
2.714
3.337
2.994
3.039
2.711
172
164
138
131
143
159
average
2.75
162
average
2.903
150
and in the way the bipyrrole fragment N7-N8-N9 interacts
with the anion (Figure 3). Specifically, the imino group N7 is
seen to participate in hydrogen binding with the phosphate oxy-
gen, but not with the bound sulfate anion; in this case, nitrogens
N8 and N9 are directed to different oxygen atoms, O(4) and
O(3). This difference is ascribed to differences in the orientation
of the tetrahedral oxyanions relative to the receptor; in 6‚H3-
PO4 the hydrogenphosphate anion is rotated by 10° in toward
O4 f O3 as compared to what is seen in the case of the
corresponding sulfate anion complex, 6H22+‚SO42- (6‚H2SO4).
7248 J. Org. Chem., Vol. 72, No. 19, 2007

Synthetic Model of the Phosphate Binding Protein
TABLE 2. Geometric Parameters Corresponding to the Anion-Stabilizing Hydrogen Bond Interactions in the SBP22 and 6‚H2SO4
SBP
6‚H2SO4
distance (Å)
sulfate
atom
distance (Å)
X‚‚‚Y
angle (deg)
X‚‚‚H‚‚‚Y
sulfate
atom
angle (deg)
X‚‚‚H‚‚‚Y
Y (atom)
Y (atom)
X‚‚‚Y
O1
OH(Ser130)
NH(Ala137)
2.76
2.70
167
151
O1
N1
N2
N6
2.986
2.740
2.982
2.839
2.923
2.938
3.349
3.420
3.347
3.042
2.640
159
148
127
166
170
171
145
159
146
141
169
O2
O3
O4
NH(Ser45)
2.84
2.83
2.77
2.82
2.67
169
170
149
157
152
O2
O3
O4
N16
N17
N9
N15
N8
N18
N20
N21
NH(Gly131)
NH(Gly132)
NH(Trp192)
NH(Asp11)
average
2.76
159
average
3.019
154
The active sites of SBP and PBP differ from one another in
terms of their hydrogen bond networks.10 Moreover, the
presence of negative and positive charges in the PBP allows it
to discriminate between phosphate and sulfate. The synthetic
system, 6H22+, bears two positively charged residues as
mentioned above. Presumably as a consequence, both the sulfate
and hydrogenphosphate anions are seen to form complexes in
the solid state. Moreover, both anions are constrained within
hydrogen bond networks that, while distinct (vide supra), are
not all that different. However, these two tetrahedral anions are
seen to bind with different affinities in acetonitrile solution and
can thus be distinguished from one another (as well from other
anions) under these ostensibly more exacting solution-phase
conditions. This latter chemistry is summarized below.
However, the yellow color of the solution of 6 become more
intense (brighter in aspect) upon the addition of excess chloride,
phosphate, or nitrate (all three anions studied in the form of
their respective tetrabutylammonium salts). These various
changes are reflected in the corresponding titration curves (cf.
Figure 4). For instance, the addition of hydrogensulfate anion
induces an increase in the intensity of the bands at 520, 530,
and 540 nm in the case of receptors 4, 5, and 6, respectively,
thereby giving rise to the observed red color. Likewise, the
addition of dihydrogenphosphate anion to all three receptors
induces an increase in the intensity at 425 nm with this increase
being characterized by a significantly larger ∆A value in the
case of 6.
-
Interestingly, upon the addition of 1 equiv of H2PO4 to 6
Anion Binding Properties of Receptors 4-6. Although the
crystal structures of 4 and 6, both as reported earlier9 and as
discussed above, provided evidence that the protonated forms
of macrocycles 4-6 are capable of interacting with anions in
the solid state, it was considered important to probe their anion
binding characteristics in solution. This had not been done
previously, even in a preliminary fashion. However, several
other smaller macrocycles were recently studied in acetonitrile.17
Therefore, for the sake of consistency, and as dictated by
solubility considerations, receptors 4-6 were analyzed for their
ability to bind various inorganic anions in this same solvent
system. As in our previous studies,8,9,17 this was done by using
standard UV-vis titrations employing the tetrabutylammonium
(TBA) salts of the anions in question and the neutral forms of
the receptors. The neutral (free base) forms were used in these
studies, rather than one or more of the protonated forms, so as
to avoid possible complications arising from incomplete pro-
tonation or counteranion effects (i.e., issues involving displace-
ment equilibria).
Prior to carrying out the actual binding studies, the UV-vis
spectra of 4 and 6 were recorded in acetonitrile. Interestingly,
in spite of the strong chemical resemblance between receptors
4 and 6, they were found to display slightly different spectra.
In both cases, three maxima were seen, with those for 6 being
shifted by 20-30 nm to longer wavelength (Figure 4). In the
case of 4, the central absorption maximum displays the highest
intensity, whereas for 6 the three absorption maxima decrease
in intensity with increasing wavelength. Of the various anions
tested, only the hydrogensulfate anion was found to engender
a significant change in color, with its addition to all three
receptors serving to change the respective acetonitrile solutions
from yellow to red. The addition of other anions does not serve
to change the color of acetonitrile solutions of either 4 or 5.
the solution becomes slightly red (Figure 5). However, the
continued addition of this anion causes the color to change back
to yellow. This observation is considered to reflect the binding
of more than one stoichiometric equivalent of dihydrogenphos-
phate anion to receptor 6. In fact, the binding stoichiometry
was determined to be 1:3 (receptor-to-anion) based on a Job
plot analysis. Figure 4 shows the changes produced upon the
-
addition of up to 1 equiv of H2PO4 and underscores the fact
the same increase in the 540 nm band intensity is seen under
these conditions as in the case of HSO4- titration. The similarity
of the UV-vis spectra for receptor 6 seen after the addition of
1 equiv of either H2PO4- or HSO4- is rationalized in terms of
the receptor folding around the anion to form a strong hydrogen
bond network similar to that observed in the solid state.
As inferred from the UV-vis titrations, all of the receptors
included in this study display a strong preference for the
hydrogensulfate and dihydrogenphosphate anions in acetonitrile
solution (>104 M-1 in all cases; approaching 107 M-1 in selected
cases). In addition receptor 6 binds chloride and acetate anions
strongly and even displays a propensity to bind large tetrahedral
anions, such as perrhenate (Ka ≈ 103 M-1; Table 3). In contrast,
receptor 4 was found to interact appreciably with only three of
the current series of test anions, namely acetate, hydrogensulfate,
and dihydrogenphosphate. Finally, receptor 5, presumably as
the result of containing more flexible methylamino linkers, can
bind chloride anion. Although receptor 6 binds a number of
anions well (i.e., with binding constants in the range 103-107
M-1), it is the hydrogensulfate anion that displays the best
combination of clean binding (i.e., 1:1 binding stoichiometry)
and high affinity (Ka ≈ 107 M-1; Table 3). In contrast,
dihydrogenphosphate is bound to this receptor well, albeit with
a 1:3 receptor-to-anion stoichiometry, as noted above. Dihy-
J. Org. Chem, Vol. 72, No. 19, 2007 7249

Katayev et al.
FIGURE 4. Changes in UV-vis spectra seen upon the addition of dihydrogenphosphate and hydrogensulfate to acetonitrile solutions of receptors
4 (left column) and 6 (right column).
TABLE 3. Anion Affinities (log Ka) for the Interactions of Test
Anions with the Bipyrrole-Derived Macrocyclic Receptors (Free
Base Forms) 4-6, As Determined from UV-Vis Spectroscopic
Titrations Carried out at 23 °C in Acetonitriled
anion/
receptor
4
5
6
a
a
a
Cl-
-
-
-
3.385(7)
4.731(7)
3.64(2)
2.20(5)
4.852(3)
3.57(8)b
3.35(8)b
6.99(1)
6.74(6); 5.98(9);
4.11(6)
a
Br-
-
-
-
a
NO3
CH3COO-
4.41(4)
3.38(3)
-
a
a
ClO4 -
-
-
a
a
ReO4 -
HSO4
-
-
4.80(3)
5.28(3);
4.38(2)
5.78(2);
5.70(3)
FIGURE 5. Colors of acetonitrile solutions of receptor 6: (1) free
- c
H2PO4
macrocycle, (2) with TBA+ Cl-, (3) with 1 equiv of TBA+ HSO4
,
-
4.78(5)
(4) with 1 equiv of TBA+ H2PO4-, and (5) with g2 equiv of
TBA+ H2PO4
.
a No apparent binding seen, as reflected in the lack of observable change
in the UV-vis spectrum upon anion addition. b Very weak changes in the
UV-vis spectrum were observed over the course of the titration; this causes
a relatively large error in derived binding constant. c Stepwise, 1:2 and 1:3
(receptor:anion) binding is observed; the values refer to those for the first,
second, and third binding events, respectively. The numbers in parentheses
are from the fits provided by the Hyperquad program; based on reproduc-
ibility, the actual experimental errors are considered to be e15%. All values
in the table are the result of at least two independent experiments. d The
corresponding tetrabutylammonium salts were used in all cases.
-
drogenphosphate also binds to receptors 4 and 5 with a 1:2
stoichiometry.
The host-guest ratio for dihydrogenphosphate binding is seen
to parallel the number of diiminobipyrrole subunits present in
the receptor and, to the extent this inference is correct, the
presence of amido-functional groups does not influence the
binding of H2PO4-. However, this conclusion does not hold
for acetate and hydrogensulfate anions, where a higher relative
affinity is observed for the pyridine-2,6-diamido containing
oriented to complement a given substrate (e.g., H2PO4-) that is
critical in terms of regulating both affinity and selectivity.
-
macrocycle 4. Moreover, for H2PO4 the association constant
for 5 is almost identical with that of 4, a finding that is consistent
with the smaller macrocycle 4 being conformationally flexible.
In other words, it appears, as expected, that it is not just the
total number of interacting groups but also the way they are
Conclusion
In this report we have detailed how large, bipyrrole-containing
receptors may be made efficiently starting from appropriate
7250 J. Org. Chem., Vol. 72, No. 19, 2007

Synthetic Model of the Phosphate Binding Protein
hydrogen atoms). The hydrogen atoms bound to nitrogen atoms
were observed in a ∆F map and refined with fixed position and
isotropic displacement parameters. Six molecules of dichlo-
romethane were found to be strongly disordered and could not be
modeled satisfactorily. The contribution to the scattering by these
molecules was removed by use of the utility SQUEEZE in
dialdehyde and diamine building blocks. This finding points
the way toward the synthesis of receptors for different anionic
species via the use of both appropriate starting materials and
anionic templates. In the present instance, the starting materials
7-9 are readily available and are expected to arrange themselves
around tetrahedral oxoanions, such as hydrogensulfate and
dihydrogenphosphate, to form receptors 4-6 after imine bond
formation. These receptors are selective for tetrabutylammonium
hydrogensulfate and tetrabutylammonium dihydrogenphosphate,
at least in acetonitrile solution. As inferred from UV-vis
spectroscopic titrations, macrocycle 6 binds dihydrogenphos-
phate anion in a stepwise 1:3 binding in acetonitrile solution
with a log Ka for the first binding event approaching 7. However,
hydrogensulfate anion is bound with a 1:1 stoichiometry with
a slightly higher affinity (log Ka even closer to 7). It is proposed
that binding of 1 equiv of HSO4 or H2PO4 to 6 induces a
conformational change that stabilizes an effective hydrogen bond
network that bears analogy to that which is seen in the solid-
state structure of the complex salt 6H22+‚HPO42-, as well as
the active center of PBP. This congruence in terms of both
structure and function leads us to suggest that receptors such
as 6 could contribute to a further understanding of the factors
that regulate anion binding in naturally occurring systems.
2
2 2
PLATON98.25 The function, ∑w(|Fo| - |Fc| ) , was minimized,
2
2
where w ) 1/[(σ(Fo))2 + (0.03P)2] and P ) (|Fo| + 2|Fc| )/3.
Rw(F2) refined to 0.1837, with R(F) equal to 0.0996 and a goodness
of fit, S, of 0.988. All calculations were carried out by use of the
SHELXTL PLUS program (PC Version 5.10).26
Tetraamine 9 (2,6-dimethylamino(2-aminophenyl)pyridine).
2,6-Pyridinedialdehyde27 (4 g, 29.60 mmol) and mono-BOC
protected o-phenylenediamine (12.4 g, 59.5 mmol) were heated at
reflux in 250 mL of MeOH for 2 h. The solution was cooled and
product was filtered off. The solids collected in this way were then
dissolved in a mixture of MeOH (200 mL) and CH2Cl2 (50 mL),
after which NaBH4 (1.44 g, 59.5 mmol) was added in small portions
over 1 h. At this point, the solution was allowed to stir overnight
at room temperature. It was then taken to dryness under reduced
pressure. The resulting solid was redissolved in a mixture of CH2-
Cl2 (100 mL) and TFA (50 mL) and stirred for 2 h. The reaction
mixture was then poured into 1 L of water containing 30 g of NaOH.
After extraction with CH2Cl2, the organic phase was dried over
MgSO4. The resulting crude product was recrystallized from
EtOAc-hexanes to give 5 g (53%) of 9 in the form of pale-yellow
crystals. MS (CI+, M+ + H) m/e calcd for C19H22N5 320.4, found
320.2. Anal. Calcd for C19H21N5: C, 71.45; H, 6.63; N, 21.93.
-
-
Experimental Section
1
Found: C, 71.11; H, 6.23; N, 21.80. H NMR (CDCl3) δ (ppm)
Titration Conditions. Stock solutions of the host molecule
subject to study were made up in acetonitrile, with the final
concentrations being in the range of 1.3-1.7 × 10-5 mol/L for
4-6. Stock solutions of the guest were prepared by dissolving 15-
25 equiv of the TBA salts of the anions in question in 5 or 10 mL
of the host stock solution (0-1 × 10-3 mol/L). Making up the
anion source solutions in this way allowed the binding studies to
be carried out without having to make mathematical corrections to
account for changes in host concentration as the result of dilution
effects. The general procedure for the UV-vis binding studies
involved making sequential additions of the titrant (anionic guest)
to a 1 mL sample of the host stock solution in the spectrometric
cell and monitoring the changes in the spectral features. The total
number of data points was between 10 and 40, depending on the
stoichiometry of complexation; for a presumed 1:1 complex 10-
15 points were usually measured. The data points were then collated
and combined to produce plots that, in turn, were processed by
HYPERQUAD.23
7.57 (t, 1H, J ) 7.6), 7.20 (d, 2H, J ) 7.6), 6.78-6.62 (m, 8H),
4.45 (s, 4H), 3.70 (br s, 4H). 13C NMR (CDCl3) δ (ppm) 157.8,
137.4, 137.2, 134.7, 134.3, 120.7, 120.2, 120.0, 118.8, 116.5, 112.2,
49.4.
Receptor 5. (Note: The full systematic name for this and other
new products can be found in the Supporting Information.) 3,3′-
Dimethyl-4,4′-dipropyl-5,5′-diformyl-2,2′-bipyrrole (7) (200 mg,
0.665 mmol) and 2,6-dimethylamino(2-aminophenyl)pyridine (9)
(212 mg, 0.665 mmol) were mixed in 5 mL of methanol, after which
2 drops of concentrated H3PO4 (about 2.5 equiv) was added. The
mixture was then stirred at rt for 48 h. The resulting dark red
precipitate was filtered and washed with methanol to the receptor
in the form of its phosphoric acid salt, 5‚(H3PO4)x (x ) 1-2
according to elemental analysis). This salt was suspended in a
mixture of dichloromethane and methanol at which point an excess
of triethylamine was added. This addition caused the solution to
turn bright yellow. After being taken to dryness in vacuo, the
resulting solid was dissolved in dichloromethane and passed through
a small alumina plug to yield 280 mg (72%) of the free receptor 5.
MS (ESI+, M+) m/e calcd for C74H82N14 1167.5, found 1168.4.
Anal. Calcd for C74H82N14: C, 76.13; H, 7.08; N, 16.80. Found:
C, 76.00; H, 7.12; N, 16.80. 1H NMR (CDCl3) δ (ppm) 9.47 (br s,
4H), 7.27 (s, 2H), 7.14 (m, 4H), 7.02 (m, 4H), 6.95 (m, 4H), 6.63
(m, 4H), 6.60 (m, 4), 4.44 (m, 8H), 2.55 (m, 12H), 2.06 (s, 8H),
1.58 (m, 8H), 0.99 (t, J ) 12 Hz, 12H). 13C NMR (CDCl3) δ (ppm)
158.5, 145.2, 142.9, 137.6, 137.1, 131.7, 128.1, 126.5, 125.7, 120.2,
119.4, 116.8, 116.3, 110.4, 49.3, 26.1, 24.7, 14.0, 10.2.
X-ray Diffraction Analysis. X-ray experimental for
[C111H113N21O6](2+)[HPO4](2-), Mr ) 1933.20: Crystals grew as
dark-red prisms and were obtained through the slow diffusion of
pentane into a dichloromethane solution of the receptor. The data
crystal was a prism that had approximate dimensions 0.24 × 0.21
× 0.18 mm3. The data were collected on a Bruker SMART 1000
CCD area detector diffractometer, using a graphite monochromator
with Mo KR radiation (λ ) 0.71073 Å). A total of 81706 reflections
were collected by using ω-scans with a scan range of 0.4 and a
counting time of 40 s per frame. The data were collected at 120 K,
using an Oxford Cryosystem low-temperature device. The crystal
is triclinic, space group P1h, a ) 21.414(4) Å, b ) 23.166(4) Å, c
) 24.601(5) Å, R ) 88.980(5)°, â ) 78.128(5)°, γ ) 84.935(5)°,
6‚H2SO4. The dihydrate of 4 (50 mg, 0.0408 mmol) prepared
according to reference 9, was dissolved in 5 mL of CH3CN at which
point TBAHSO4 (14 mg, 0.0412 mmol) was added. The solution
was then left without stirring for 5 days before being layered with
50 mL of ether. The resulting precipitate was then filtered off and
dried in vacuo to yield 25 mg (47%) of 6‚H2SO4 in the form of red
Z ) 4, V ) 11896(4) Å,3 Fcalc ) 1.079 g/cm3, µ ) 0.084 mm-1
,
θmin ) 0.88°, θmax ) 25.09°. Data reduction was performed with
SAINTPlus.24 The structure was solved by direct methods and
refined by full-matrix least-squares on F2 with anisotropic displace-
ment parameters for the non-H atoms. The hydrogen atoms on
carbon were calculated in ideal positions with isotropic displacement
parameters set to 1.2Ueq of the attached atom (1.5Ueq for methyl
(25) Spek, A. L. In PLATON, A Multipurpose Crystallographic Tool;
Utrecht University: Utrecht, The Netherlands, 1998.
(26) Sheldrick, G. M. In SHELXTL, v. 5.10, Structure Determination
Software Suite; Bruker AXS: Madison, WI, 1998.
(23) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739-1753.
(24) SAINTPlus, B; Bruker AXS: Madison, WI, 1998.
(27) Katayev, E. A.; Reshetova, M. D.; Ustynyuk, Y. A. Russ. Chem.
Bull. Int. Ed. 2004, 53, 335-339.
J. Org. Chem, Vol. 72, No. 19, 2007 7251

Katayev et al.
crystals. MS (ESI+, M+) m/e calcd for C111H111N21O6 1835.2, found
11.69 (br s, 2H), 10.62 (m, 5H), 10.36 (m, 2H), 8.49 (m, 2H), 8.37
(m, 4H), 8.06 (m, 5H), 7.64 (m, 4H), 7.47 (m, 2H), 7.37 (s, 2H),
7.09 (m, 2H), 7.01 (m, 2H), 6.84 (m, 7H), 6.70 (m, 4H), 6.49 (m,
2H), 6.04 (m, 4H), 2.94 (m, 2H), 2.62 (m, 4H), 2.24 (m, 3H), 2.08
(m, 4H), 2.03 (s, 6H), 1.93 (s, 6H), 1.90 (s, 6H), 1.64 (m, 5H),
1.55 (m, 4H), 1.24 (m, 5H), 1.03 (m, 12H), 0.49 (t, J ) 7.2, 6H).
13C NMR (CDCl3) δ (ppm) 163.8, 160.9, 160.2, 152.1, 150.1, 149.5,
148.7, 148.5, 145.8, 143.7, 139.2, 138.8, 138.3, 137.6, 136.8, 136.4,
132.6, 131.5, 130.6, 130.4, 129.8, 127.9, 127.09, 126.8, 125.6,
124.9, 124.6, 124.5, 124.4, 124.1, 123.7, 123.2, 122.7, 122.2, 121.6,
119.9, 119.4, 118.5, 118.1, 114.3, 114.3, 99.9, 27.1, 26.9, 26.3,
24.9, 24.7, 22.8, 13.9, 13.9, 13.8, 13.6, 13.0, 12.6.
1835.0. Anal. Calcd for C111H113N21O10S: C, 68.96; H, 5.89; N,
1
15.21. Found: C, 68.93; H, 5.95; N, 15.20. H NMR (CDCl3) δ
(ppm) 12.42 (s, 2H), 12.00 (m, 2H), 10.95 (s, 2H), 10.70 (s, 2H),
10.31 (s, 2H), 9.94 (s, 2H), 9.67 (s, 2H), 8.44 (t, J ) 8.4 Hz, 4H),
8.40 (d, 2H), 8,13 (m, 3H), 8.02 (d, J ) 7.2 Hz, 2H), 7.60 (d, J )
7.8 Hz, 2H), 7.45(m, 4H), 7.39 (m, 2H), 7.04 (m, 4H), 6.91 (m,
6H), 6.78 (m, 2H), 6.71 (d, J ) 7.2 Hz, 2H), 6.50 (d, J ) 7.8 Hz,
2H), 6.20 (m, 4H), 2.95 (m, 2), 2.60 (m, 4H), 2.14 (m, 6H), 1.99
(s, 6H), 1.80 (s, 14H), 1.65 (m, 5H), 1.51 (m, 5H), 1.26 (m, 5H),
1.00 (t, J ) 7.2 Hz, 6H), 1.03 (t, J ) 7.2 Hz, 6H), 0.49 (t, J ) 7.2
Hz, 6H). 13C NMR (CDCl3) δ (ppm) 163.8, 161.1, 161.0, 150.7,
150.3, 150.2, 149.2, 148.9, 148.4, 146.6, 142.4, 138.8, 137.8, 137.1,
134.3, 132.8, 131.5, 130.7, 130.5, 129.5, 127.9, 127.5, 127.3, 127.2,
126.6, 125.9, 124.8, 124.7, 124.6, 124.6, 124.2, 123.8, 123.5, 122.8,
122.8, 121.6, 119.6, 119.6, 118.9, 118.7, 118.7, 118.2, 114.2, 99.9,
99.9, 98.3, 27.1, 26.6, 26.2, 24.8, 24.6, 22.9, 13.9, 13.9, 13.7, 13.2,
12.8, 12.4.
Receptor 6. Free receptor 6 was obtained from salt 6‚H2SO4 or
6‚H3PO4 by mixing it with an excess of triethylamine in dichlo-
romethane, passing the resulting mixture through a small plug of
silica gel, and evaporating the solution that elutes to dryness at rt.
Acknowledgment. This work was supported in part by the
U.S. National Institutes of Health (grant no. GM 58907 to J.L.S.)
and the Russian Foundation for basic research (grant nos. 05-
03-32684 and 05-03-08017).
6‚H3PO4. 3,3′-Dimethyl-4,4′-dipropyl-5,5′-diformyl-2,2′-bipyr-
role9 (7) (40 mg, 0.133 mmol) and bis(2-aminophenyl)pyridine-
2,6-dicarboxamide (8) (42 mg, 0.144 mmol) were mixed in 1 mL
of methanol after which 2 drops of concentrated H3PO4 (about 2.5
equiv) was added. The mixture was stirred for 48 h at rt. The
resulting orange precipitate was filtered off and washed with
methanol to yield 80 mg (46%) of the salt 6‚H3PO4. MS (ESI+,
M+) m/e calcd for C111H111N21O6 1835.2, found 1835.0. Anal. Calcd
for C111H114N21O10P: C, 68.96; H, 5.94; N, 15.22. Found: C, 68.95;
Supporting Information Available: Experimental part, UV-
vis titrations curves, as well as X-ray crystallographic information
including CIF files. This material is available free of charge via
the Internet at http://pubs.acs.org.
1
H, 5.87; N, 15.22. H NMR (CDCl3) δ (ppm), 14.30 (br s, 2H),
JO071106G
7252 J. Org. Chem., Vol. 72, No. 19, 2007