Exploring the effects of cooperative interactions on affinity using a pinwheel sensor system

Article abstract of DOI:10.1016/j.tet.2004.08.048

A series of three pinwheel sensors were constructed with 1, 2, and 3 binding sites. Binding of Zn⁺² and Cd⁺² was monitored by fluorescence over a range of temperatures. The data demonstrate that cooperative interactions generally increase the effective affinity of the sensor. This effect is more pronounced in systems which have lower inherent affinity for the analyte.

Full text of DOI:10.1016/j.tet.2004.08.048

Tetrahedron 60 (2004) 11057–11065
Exploring the effects of cooperative interactions on affinity using a
pinwheel sensor system
*
Paul D. Jones and Timothy E. Glass
Department of Chemistry, The Pennsylvania State University, University Park, PA 16802, USA
Received 5 February 2004; revised 7 July 2004; accepted 19 August 2004
Available online 21 September 2004
Abstract—A series of three pinwheel sensors were constructed with 1, 2, and 3 binding sites. Binding of ZnC2 and CdC2 was monitored by
fluorescence over a range of temperatures. The data demonstrate that cooperative interactions generally increase the effective affinity of the
sensor. This effect is more pronounced in systems which have lower inherent affinity for the analyte.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
challenge of receptor design becomes one of creating the
optimal spacing and orientation of the recognition elements
in a rigid fashion.8 For receptors of the class depicted in
Scheme 1, there is a range of distances which the
recognition elements can span owing to free rotation
about the acetylenic axis. Binding of the first analyte rigidly
sets the correct distance between the remaining pairs of
binding groups. Thus, the entropic loss upon binding the
analyte is effectively averaged over three consecutive
binding events, diminishing the overall entropic penalty to
binding. This receptor design should be general for a range
of analytes of varying size within the outer limits of receptor
framework. The spacing between the recognition elements
need only be roughly adjusted to the size of the desired
guest. Upon binding of the first guest, the receptor is then
fully organized to match the size of the guest.
Fluorescent chemical sensors are becoming an increasingly
valuable tool for the detection of many analytes in a wide
range of applications.1 As part of an ongoing project geared
toward creating more effective chemical sensors, we have
been exploring a novel class of cooperative receptors based
on bistrityl acetylene compounds, termed pinwheel recep-
tors (Scheme 1).2 The recognition elements appended to the
trityl skeleton bind the analyte across the receptor frame-
work producing three identical binding pockets. Coopera-
tive recognition is seen because binding of the first analyte
preorganizes the symmetrical receptor into a conformation
in which the second and third analyte can bind more
strongly. This mechanism of cooperativity is conceptually
similar to the restricted rotational freedom employed by
Rebek’s cooperative biphenyl bis-crown ether3 and
Shinkai’s cooperative porphyrin sandwich complexes.4
We have previously argued that the type of cooperative
recognition described above increases the affinity and
selectivity of a sensor for its analyte relative to a similar
non-cooperative sensor.2a These points were initially
addressed using sensors for metal ions2b and dicarboxyl-
ates.9 In this report, we explore these issues using a series of
In fact, several artificial cooperative receptors have been
developed,5 though few have been applied toward chemical
sensor applications.6 From a design perspective, this method
of recognizing a guest is attractive. Many elegant (non-
cooperative) receptor designs have been developed over the
years in which two or more recognition elements are
displayed in a convergent fashion for binding of a guest.
Strong binding is most often observed when the recognition
elements are held rigidly apart such that there is little
entropic loss to the receptor upon binding the guest.7 The
Keywords: Sensor; Allosteric; Cooperative.
* Corresponding author. Address: Department of Chemistry, University of
Missouri, Columbia, MO 65211, USA. Tel.: C1-573-882-3813; fax: C1-
573-882-2754; e-mail: glasst@missouri.edu
Scheme 1.
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.08.048

11058
P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
metal ion sensors with varying numbers of binding sites.
Comparison of the recognition properties of sensors with
varying degrees of cooperativity provides insight into the
advantages and disadvantages of the cooperative recog-
nition method.
alkynes was effected by copper chloride and N-methyl
pyrrolidine or TMEDA under aerobic conditions. Treatment
of butadiyne 8a–c with excess titanium tetrachloride and
a,a-dichloromethyl methyl ether regioselectively formyl-
ated only the anisole rings, yielding aldehydes 9a–c.
Reductive amination of the aldehydes with amine 5 using
triacetoxy sodium borohydride yielded sensors 2a–c.
2. Results
2.2. Metal ion binding studies
The sensors used for this study are shown in Figure 1.
Compounds 2a–c are based on a bis-acetylene bis-trityl
system for ease of synthesis.10 The amino-methoxyquinal-
dine groups would serve not only as recognition elements,
but also as the fluorescent readout mechanism. Two
recognition elements (R) should bind a metal ion between
them, across the bisacetylene axis in a tetrahedral fashion
(as per Fig. 1). The benzylic amine will participate in
photoelectron transfer (PET)1b quenching of the fluorophore
in the unbound state. Upon metal chelation the amine will
no longer be able to quench the fluorophore and a significant
increase in fluorescence is anticipated. Receptor 2a, with
one set of binding groups, was designed to be a non-
cooperative receptor while receptors 2b and 2c were
designed to be two-fold and three-fold cooperative recep-
tors, respectively. Assuming that the binding pockets in the
three molecules are similar, this series of sensors should
provide information on the effect of cooperativity on the
recognition properties of the system.
2.2.1. UV/vis studies. Metal ion binding was first studied by
titrating sensors 2a–c with metal ions and following changes
in the UV/vis absorptions. All three sensors appeared to
bind many metals tightly including NiC2, CdC2, ZnC2
,
MnC2, HgC2, and AgC1. Figure 2 shows a representative
example of the titrations of sensors 2a–c with Cd(ClO4)2 in
acetonitrile. All metals produced a red-shift in the major UV
band centered at 325 nm. With the exception of extinction
coefficient, which varied as per the number of chromo-
phores per molecule, all three sensors had similar UV
behavior indicating that the binding pockets are similar
between the various sensors.
Unfortunately, all of the binding constants were too high to
derive useful data from the UV/vis titrations. In fact,
titration curves based on Figure 2 showed near linear
responses which saturated at approximately the stoichio-
metric point for each sensor (1, 2, and 3 equiv of CdC2 for
sensors 2a, 2b, and 2c, respectively). These results verify
the expected stoichiometry of the sensors.
2.1. Synthesis
The synthesis of receptors 2a–c is shown in Scheme 2. The
fluorophore portion was prepared by bromination of
methoxyquinaldine followed by substitution with methyl-
amine to give 5. Trityl chlorides 6a–c were reacted with
ethynyl Grignard to yield alkynes 7a–c. Dimerization of the
2.2.2. Fluorescence studies. The fluorescence titrations of
the sensors were much more informative as the concen-
tration of sensor required for such analysis was significantly
lower (0.3 mM). Of all of the metals tested, only zinc and
cadmium produced well behaved fluorescence modulation
Figure 1. A series of three bis-trityl receptors.

P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
11059
Scheme 2.
when added to the sensor. Figure 3 shows a representative
example in which 2a is titrated with ZnC2 in acetonitrile. As
anticipated, the sensor has little native fluorescence due to
the PET quenching of the benzylic amines. Binding of the
metal ions produced a substantial increase in fluorescence.
Generally, cadmium produced a 3–4 fold increase in
emission at 393 nm (excited at 336 nm) and zinc produced
a 7–9 fold increase in emission at the same wavelength. It is
likely that quenching metal ions such as silver did not give a
good response with this sensor because they quench the
fluorophore as they bind, retaining an ‘off’ state of the
sensor.
of free guest in solution very well. Since the binding
isotherms used to calculate the dissociation constants rely
on free guest concentrations, the data were corrected using
the known concentration of sensor so that the x-axis reflects
the equilibrium concentration of free zinc in solution. Using
this correction, more accurate dissociation constants and
Hill coefficients could be obtained. Therefore, this correc-
tion was applied to all titration data.
Binding constants for the three sensors with ZnC2 and CdC2
were determined by fitting the experimental data to the Hill
equation (Table 1). For the purpose of internal consistency,
sensor 2a was fitted to the Hill equation and in all cases gave
a Hill coefficient of very near 1.0, which is consistent with a
non-cooperative sensor.
The binding isotherms for all three sensors with zinc are
shown in Figure 4. While the fluorescence experiment
allows the concentration of host to be very low, the apparent
dissociation constants of the receptors are still close to the
concentration of host. This implies that the concentration of
added guest does not really approximate the concentration
The UV/vis titration data was used to determine the
stoichiometry of the sensors. In order to verify this in the
fluorescence mode, job plots were prepared. As expected,

11060
P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
Figure 3. Fluorescence titration of compound 2a with Zn(ClO4)2 at 25 8C.
[2a]Z0.3 mM in CH3CN with 0.50 mM NaClO4 as a supporting electrolyte.
lexZ336 nm.
Figure 4. Binding isotherms for receptors 2a–c with Zn(ClO4)2 at 393 nm
in acetonitrile with 0.50 mM NaClO4 as a supporting electrolyte.
Concentration of ZnC2 is given as the equilibrium concentration of free
zinc in solution. (A) Receptor 2a. The best fit line (solid) is to a single site
isotherm. (B) Receptor 2b. The best fit line (short dash) is to a Hill equation.
(C) Receptor 2c. The best fit line (long dash) is to a Hill equation.
Zn(ClO4)2 at different temperatures to investigate the effect
of cooperativity on apparent Kd. Figure 6 summarizes these
experiments. Generally, the Kd of the non-cooperative
sensor increased with increasing temperature, while the Kd
of the cooperative sensors was relatively unaffected.
Figure 2. UV/Vis titration of sensors 2a–c ([2]Z10 mM) with Cd(ClO4)2 in
CH3CN with 0.5 mM NaClO4 as a supporting electrolyte. (A) Sensor 2a.
(B) Sensor 2b. (C) Sensor 2c.
each sensor gave a maximum change in fluorescence close
to its stoichiometric point for both metals tested. The Job
plot for the three sensors and CdC2 are overlaid in Figure 5.
Sensor 2a had maximal change in fluorescence at 48 mol%
fraction of CdC2, sensor 2b at 67% and sensor 2c at 73%.
Interestingly, sensors 2b and 2c show suppressed changes in
fluorescence at low mole fraction of metal, in keeping with
the cooperative nature of these receptors.
3. Discussion
Shinkai has described a detailed study of five receptors
based on his cooperative cerium sandwich complexes with
1, 2, 3, and 4 sites.11 While the Hill coefficient correlated
nicely with the number of sites, the binding pockets
communicated with each other electronically such that it
was not possible to compare association constants in a
meaningful way. In this study, we attempted to make
comparisons of dissociation constants between sensors with
2.2.3. Variable temperature studies. Fluorescence
titrations were then conducted with receptors 2a–c and

P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
11061
Table 1. Hill coefficients and dissociation constants for sensors 2a–c under conditions listed in Figure 4
Sensor
Analyte
Hill coeff.
Kd (mM)a
I/I0 max.b
2a
2b
2c
2a
2b
2c
ZnC2
ZnC2
ZnC2
CdC2
CdC2
CdC2
0.97
1.9
2.9
0.99
2.0
2.9
0.85
0.54
0.33
0.42
0.26
0.26
9.2
7.3
7.2
4.8
4.7
7.6
a
Error in Kd is approximately G15%.
The maximum fluorescence change of each sensor at saturation.
b
different numbers of binding pockets. It has been shown
previously for a related pinwheel sensor, that two sets of
binding groups on one trityl unit cannot interact to bind a
metal, therefore binding must occur across the acetylene
axis as shown in Figure 1. In the case of sensors 2a–c, the
similarity of the three sets of UV titration data indicates that
the binding sites are similar between the three sensors. This
result is supported by NMR titrations of the three sensors
with metal ions (data not shown) in which similar behavior
is observed between all three sensors. Therefore, to a first
approximation, the binding pocket in 2a appears to be
equivalent to those of 2b and 2c.
sensor. Moreover, the 3-fold cooperative system (2c) has a
lower apparent Kd than the 2-fold cooperative sensor (2b).
Similarly, the effective range of analyte over which the
sensor is useful is larger for the non-cooperative system (2a)
than 2b which is again larger than the 2c. Thus, the
advantages and disadvantages of cooperativity are immedi-
ately evident.
As expected, the hill coefficients for sensors 2b and 2c
correlated to the number of binding sites indicating that the
multi-site sensors are binding in a cooperative mode. It is
evident from Figure 4 and Table 1 that for zinc, the
cooperative sensors bind more tightly than the non-
cooperative sensors. However, the differences in this case
(ca. 2.5-fold decrease in Kd for 2c vs. 2a) are small
compared to what we have observed previously (50-fold).2a
We believe that the difference arises from the high inherent
affinity of 2 for metal ions. The contribution of cooperativity
arises from the freezing out of rotational entropy upon
binding of the first metal ion. The higher affinity of the
second metal ion can be viewed in terms of free energy of
association in which the total free energy is a combination
of the inherent free energy of the recognition elements for
the metal plus the free energy of cooperativity12 (which
stems from preorganization of the sensor upon the first
binding event). Therefore, it stands to reason that in this
system where the inherent affinity of the sensor for zinc is
quite high (large free energy of association) the added free
energy of cooperativity makes less of an impact in terms of
higher affinity compared to a system with lower inherent
affinity for the analyte. This assertion is qualitatively
supported by the cadmium binding data (Table 1) in
which 2a binds CdC2 even more tightly than ZnC2 and
To compare affinity between sensors, binding constants
were determined in terms of apparent dissociation constants
(Kd). This value is the half saturation value of the sensor, a
value that is commonly used in biological systems when the
number of binding sites is unknown.12 In this case, it allows
direct comparison of affinity between sensors with different
numbers of sites, since the actual association constants of
the three sensors have different units. This type of analysis is
also convenient in cases where the Hill coefficient is non-
integral implying a mixture of species of varying analyte
occupancy. Ultimately, this value is the most useful measure
of binding as it describes the concentration of analyte that
the sensor is capable of recognizing.
Visual inspection of the binding isotherms in Figure 4
supports our assertion that cooperativity enhances binding.
The cooperative sensors have steeper binding isotherms and
saturate much more quickly giving an overall lower
apparent Kd (higher binding) than the non-cooperative
Figure 5. Job plot for receptors 2a–c with Cd(ClO4)2. [Cd2C]C[host]Z
20 mM in CH3CN with 0.50 mM NaClO4 as a supporting electrolyte. The
lines are merely illustrative and do not represent a fit to the data.
Figure 6. Apparent Kd of sensors 2a–c with Zn(ClO4)2 as a function of
temperature.

11062
P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
the ratio of Kd for 2c and 2a is smaller. Thus, it is reasonable
to expect that cooperativity will have the biggest impact on
association in systems with relatively low inherent free
energy of association.
the experimentally determined binding isotherms to the Hill
equation using Kaleidagraph or by the standard linearization
techniques for cooperative systems.13
5.2. Synthesis
To explore these effects further, we examined the effect of
temperature on Kd. The plot of Kd versus temperature
(Fig. 6) has some scatter due to the error inherent in
calculation of Kd, however, the trends are very informative.
The non-cooperative sensor demonstrated the greatest
variability in Kd over the relatively small range of
temperatures explored. The increase in Kd (decrease in
affinity) as temperature increased is in agreement with
expectation since the organization of the sensor to bind the
metal is entropically disfavored. At higher temperatures,
binding affinity should be decreased. Interestingly, the
cooperative sensors showed little variation in apparent Kd
over the temperature range. Again this is consistent with the
picture of a sensor architecture in which the entropic loss in
organizing the system for binding is averaged over multiple
binding events. Thus the binding of the cooperative
receptors is expected to be dominated by the enthalpic
contribution of the metal–ligand bonds. It is evident that as
the binding affinity of the non-cooperative system decreases
(at higher temperature), the effect of cooperativity is more
pronounced. Thus the cooperative sensor 2c has a five-fold
greater effective binding constant for zinc than 2a at the
higher temperatures.
5.2.1. 2-Bromomethyl-6-methoxy-quinoline 4. A mixture
of 6-methoxyquinaldine (820 mg, 4.7 mmol) and NBS
(838 mg, 4.7 mmol) was refluxed in CCl4 (100 mL).
AIBN (40 mg, 0.24 mmol) was added to the solution. The
reaction was refluxed for an additional 3 h then cooled to
10 8C. The reaction mixture was filtered and the filtrate was
concentrated in vacuo. The resultant residue was purified
via chromatography (SiO2 1:1 DCM–hexanes) to give
compound 4 as a pink solid (732 mg, 62%). MpZ98–
100 8C dec. 1H NMR (300 MHz, CDCl3): 8.05 (d, JZ8.5 Hz,
1H), 7.96 (d, JZ9.2 Hz, 1H), 7.52 (d, JZ8.5 Hz, 1H), 7.37
(dd, JZ9.2, 2.8 Hz, 1H), 7.12 (d, JZ2.8 Hz, 1H), 4.69 (s,
2H), 3.93 (s, 3H). 13C NMR (50 MHz, CDCl3): 159.2,
155.3, 144.5, 136.9, 131.5, 129.2, 123.5, 122.3, 105.7, 55.9,
34.8. FTIR (neat): 2938, 1625, 1599, 1503, 1482, 1380,
1254, 1225, 1163, 1117, 1030 cmK1. HRMS (m/z):
calculated for C11H11BrNO [MCH]C: 252.0019, found:
252.0017.
5.2.2. (6-Methoxy-quinolin-2-yl-methyl)-methylamine 5.
Bromide 4 (362 mg, 1.4 mmol) was dissolved in THF
(2.0 mL) and added via syringe to a solution of MeNH2 in
EtOH (24.2 g, 5.7 mmol MeNH2/g solution) over 1 hr at
room temperature. It was found that using the commercially
available solutions of methylamine (40% in water or 1 M in
THF) gave poorer results. The reaction was stirred for an
additional 2 h. The solvent was removed and DCM (50 mL)
was added to the resultant solid. Aqueous NaOH (2.0 M)
was added via syringe to the stirred DCM suspension until
all solids were dissolved (ca. 800 mL). The organic layer
was separated and dried with MgSO4. The solvent was
removed and the remaining oil purified by chromatography
(SiO2, 15% TEA in EtOAc) to give compound 5 as a yellow
4. Conclusions
A series of three sensors for metal ions has been synthesized
which vary by the number of binding sites. The sensors
incorporate an amino-methoxyquinaldine moiety which
functions not only as a recognition unit but also as a PET
based fluorescent readout. Fluorescence studies were
performed at various temperatures and association was
measured in terms of apparent Kd. Binding data for zinc and
cadmium ions indicated that cooperative interactions
produced an increase in affinity by lowering the entropic
penalty for organizing the receptor for tight binding. The
extent of the increase in affinity depended on the inherent
strength of the recognition elements for the analyte. Taken
together, these studies highlight the advantages of coopera-
tive recognition.
1
oil (262 mg, 93%). H NMR (360 MHz, CDCl3): 7.99 (d,
JZ8.5 Hz, 1H), 7.95 (d, JZ9.3 Hz, 1H), 7.39 (d, JZ8.5 Hz,
1H), 7.34 (dd, JZ9.2, 2.8 Hz, 1H), 7.05 (d, JZ2.8 Hz, 1H),
4.01 (s, 2H), 3.90 (s, 3H), 2.53 (s, 3H), 2.26 (br, 1H). 13C
NMR (90 MHz, CDCl3): 157.6, 157.4, 143.8, 135.1, 130.3,
128.1, 121.9, 120.8, 105.1, 57.7, 55.4, 36.3. FTIR (neat):
3321, 2937, 2837, 1622, 1601, 1568, 1501, 1461, 1380,
1233, 1163, 1111, 1030 cmK1. HRMS (m/z): calculated for
C12H15N2O [MCH]C: 203.1179, found: 203.1189.
5. Experimental
5.2.3. 4-Methoxytrityl alkyne 7a. Chloride 6a (2.5 g,
8.1 mmol) was dissolved in argon-sparged toluene
(300 mL) and cooled to K78 8C. To this was added
ethynylmagnesium bromide (32 mL, 0.5 M in THF) via
syringe over 10 min. The reaction was allowed to warm to
room temperature and was stirred an additional 3 h. The
reaction was quenched with saturated aqueous NH4Cl and
extracted with EtOAc. The organic layer was dried with
MgSO4, filtered, and concentrated in vacuo. The residue
was purified via chromatography (SiO2, 1:5 DCM–hexanes)
to give compound 7a as a white solid (1.889 g, 78%). MpZ
5.1. General procedures
NMR spectra were recorded on a Bruker WP-200, AC-200,
DPX-300, AMX-360, DRX-400 in CDCl3 with residual
protonated solvent or tetramethylsilane (TMS) as an internal
reference unless otherwise noted. Fluorescence titrations
were conducted on a Shimadzhu 5301PC spectrofluorimeter
using 0.5 or 1.0 mL sample volumes with 0.3 mM sensor and
0.5 mM NaClO4 as a supporting electrolyte. UV/vis
titrations were conducted on a Cary 1E spectrophotometer
using 1.0 mL sample volumes and 10 mM sensor and
0.5 mM NaClO4 as a supporting electrolyte. Dissociation
constants and Hill coefficients were determined by fitting
1
100–103 8C. H NMR (300 MHz, CDCl3): 7.29–7.15 (m,
10H), 7.17–7.14 (m, 2H), 6.82–6.79 (m, 2H), 3.77 (s, 3H),
2.67 (s, 1H). 13C NMR (75 MHz, CDCl3): 158.4, 145.0,

P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
11063
136.9, 130.1, 129.0, 127.9, 126.8, 113.3, 90.0, 73.2, 55.2,
54.8. FTIR (neat): 3301, 3054, 2986, 2306, 1606, 1509,
1490, 1445, 1421, 1264, 1180 cmK1. HRMS (m/z):
calculated for C23H19O [MCH]C: 299.1430, found:
299.1424.
7.45 (d, JZ2.5 Hz, 2H), 7.32 (d, JZ8.5 Hz, 2H), 7.30–7.20
(m, 22H), 6.98 (d, JZ2.7 Hz, 2H), 6.97 (dd, JZ8.6, 2.5 Hz,
2H), 6.78 (d, JZ8.7 Hz, 2H), 3.83 (s, 6H), 3.73 (s, 6H), 3.69
(s, 4H), 3.55 (s, 4H), 2.14 (s, 6H). 13C NMR (75 MHz,
CDCl3): 158.3, 157.2, 156.6, 144.8, 143.3, 136.1, 135.1,
131.1, 130.2, 129.0, 128.4, 128.2, 128.0, 126.9, 126.6,
121.5, 121.3, 109.6, 105.2, 84.5, 69.8, 64.1, 55.7, 55.5, 55.3,
42.5. FTIR (neat): 2939, 2836, 1623, 1601, 1558, 1498,
1456, 1377, 1310, 1234, 1161, 1110, 1031 cmK1. HRMS
(m/z): calculated for C70H64N4O4 [MC2H]2C: 512.2458,
found: 512.2439.
5.2.4. 1,4-Bis-(4-methoxytrityl)-butadiyne 8a. Alkyne 7a
(1.02 g, 3.40 mmol) and CuCl (3.36 g, 34.0 mmol) were
added to DCM (35 mL) at room temperature. N-methyl
pyrrolidine (7.0 mL, 68 mmol) was added drop wise to the
reaction. O2 was bubbled through the reaction via a
dispersion tube for 8 h. The reaction was filtered through
a SiO2 plug with EtOAc to remove the copper salts. The
filtrate was concentrated and the residue purified via
chromatography (SiO2, 15% v/v DCM in hexanes) to give
compound 8a as an off white solid (626.7 mg, 61%). MpZ
5.2.7. 4,40-Dimethoxytrityl-alkyne 7b. Chloride 6b (2.5 g,
7.4 mmol) was dissolved in argon sparged toluene (300 mL)
and cooled to K78 8C. To this was added ethynylmagne-
sium bromide (20 mL, 0.5 M in THF) via syringe over
10 min. The reaction was allowed to warm to room
temperature and was stirred an additional 3 h. The reaction
was quenched with saturated aqueous NH4Cl and extracted
with EtOAc. The organic layer was dried with MgSO4,
filtered, and concentrated in vacuo. The residue was purified
via chromatography (SiO2, 1:4 DCM–hexanes) to give
compound 7b as a white solid (1.400 g, 57%). MpZ107–
1
187–188 8C. H NMR (360 MHz, CD2Cl2): 7.32–7.23 (m,
20H), 7.15–7.13 (m, 4H), 6.86–6.82 (m, 4H), 3.78 (s, 6H).
13C NMR (75 MHz, CD2Cl2): 159.1, 145.0, 136.7, 130.5,
129.3, 128.5, 127.4, 113.8, 84.9, 70.0, 56.0, 55.6. FTIR
(neat): 3057, 2954, 2835, 2037, 1954, 1901, 1605, 1582,
1508, 1489, 1462, 1445, 1298, 1251, 1178 cmK1. HRMS
(m/z): calculated for C44H35O2 [M]C: 594.2553, found:
594.2564.
1
108 8C. H NMR (300 MHz, CDCl3): 7.35–7.25 (m, 5H),
7.22–7.19 (m, 4H), 6.87–6.84 (m, 4H), 3.83 (s, 6H), 2.72 (s,
1H). 13C NMR (75 MHz, CDCl3): 158.3, 145.3, 137.2,
130.1, 128.9, 128.0, 126.8, 113.3, 90.2, 73.0, 55.2, 54.1.
FTIR (neat): 3053, 2986, 2305, 1607, 1509, 1421, 1422,
1265, 1179, 1034 cmK1. HRMS (m/z): calculated for
C23H21O2 [MCH]C: 329.1536, found: 329.1515.
5.2.5. 1,4-Bis-(4-methoxytrityl-3-carboxaldehyde)-buta-
diyne 9a. Bis-trityl 8a (280 mg, 0.47 mmol) and a,a-
dichloromethyl methyl ether (460 mL, 5.0 mmol) were
dissolved in DCM (60 mL) and the resulting solution
cooled to K5 8C. TiCl4 (660 mL, 6.0 mmol) was added drop
wise via syringe over 2 min. The reaction was allowed to
warm to room temperature and was stirred an additional
1 hr. The reaction was poured into aqueous HCl (100 mL,
2.0 M) and stirred for 2 min. The aqueous suspension was
extracted with EtOAc (3!100 mL). The organic extracts
were combined, dried with MgSO4 and concentrated. The
resultant residue was purified via column chromatography
(SiO2, 1:1 DCM–EtOAc) to give compound 9a as an off
white amorphous solid (300 mg, 98%). 1H NMR (300 MHz,
CD2Cl2): 10.4 (s, 2H), 7.56 (dd, JZ8.7, 2.7 Hz, 2H), 7.51
(d, JZ2.5 Hz, 2H), 7.35–7.20 (m, 20H), 7.01 (d, JZ8.7 Hz,
2H), 3.93 (s, 6H). 13C NMR (75 MHz, CD2Cl2): 189.4,
161.4, 144.2, 137.1, 136.8, 129.2, 128.7, 128.6, 127.7,
124.5, 112.3, 84.6, 70.4, 56.3, 56.0. FTIR (neat): 3058,
2942, 2861, 1684, 1604, 1579, 1490, 1446, 1417, 1393,
1282, 1254, 1205, 1180, 1110, 1025 cmK1. HRMS (m/z):
calculated for C46H35O4 [MCH]C: 651.2530, found:
651.2555.
5.2.8. 1,4-Bis-(4,40-dimethoxytrityl)-butadiyne 8b.
Alkyne 7b (691 mg, 2.1 mmol) and CuCl (2.0 g, 21 mmol)
were added to DCM (20 mL) at room temperature. N-
methyl pyrrolidine (4.4 mL, 42 mmol) was added drop wise
to the reaction. O2 was bubbled through the reaction via a
dispersion tube for 8 h. The reaction was poured into
100 mL dilute aqueous HCl and extracted with DCM. The
organic layer was dried with MgSO4, concentrated in vacuo
and the residue purified via chromatography (SiO2, 20% v/v
DCM in hexanes) to give compound 8b as an off white solid
(420 mg, 64%). MpZ172–173 8C. 1H NMR (360 MHz,
CD2Cl2): 7.35–7.23 (m, 10H), 7.18–7.16 (m, 4H), 6.86–6.84
(m, 4H), 3.80 (s, 6H). 13C NMR (75 MHz, CD2Cl2): 159.1,
145.4, 137.1, 130.5, 129.3, 128.5, 127.3, 113.8, 85.2, 69.9,
55.6, 55.4. FTIR (neat): 3055, 2955, 2836, 2042, 1960,
1894, 1734, 1606, 1583, 1507, 1464, 1444, 1415, 1300,
1251, 1177, 1033 cmK1. HRMS (m/z): calculated for
C46H39O4 [MCH]C: 655.2843, found: 655.2831.
5.2.6. 1,1,4,4-Tetraphenyl-1,4-bis(2-methoxy-benz-4-yl-
(6-methoxyquinolin-2-ylmethyl) methylamino)-buta-
diyne 2a. Aldehyde 9a (60 mg, 0.077 mmol) and amine 5
(60 mg, 0.30 mmol) were dissolved in DCM (3.0 mL). To
this was added HOAc (14 mL, 0.22 mmol) and
NaBH(OAc)3 (61 mg, 0.30 mmol). The reaction was
allowed to stir for 10 h. Aqueous NaOH (1 M, 2 mL) was
added and the reaction was stirred for 5 min. The organic
layer was separated and dried with MgSO4. The solvent was
removed and the resultant residue was purified via
chromatography twice (SiO2, 2% TEA/EtOAc then 2%
MeOH(NH3)/DCM) to give compound 2a as an amorphous
5.2.9. 1,4-Bis-(4,40-dimethoxytrityl-3,30-di-carboxalde-
hyde)-butadiyne 9b. Bis-trityl 8b (420 mg, 0.64 mmol)
and a,a-dichloromethyl methyl ether (574 mL, 6.4 mmol)
were dissolved in DCM (60 mL) and the resulting solution
cooled to K5 8C. TiCl4 (840 mL, 7.6 mmol) was added drop
wise via syringe over 2 min. The reaction was allowed to
warm to room temperature and was stirred an additional
20 min. The reaction was poured into aqueous HCl (100 mL,
2.0 M) and stirred for 2 min. The aqueous suspension was
extracted with EtOAc (3!100 mL). The organic extracts
were combined, dried with MgSO4 and concentrated. The
resultant residue was purified via chromatography (SiO2,
1:1 DCM–EtOAc) to give compound 9b as a pink
1
yellow solid (53 mg, 66%). H NMR (300 MHz, CD3CN/
CD2Cl2): 7.82 (d, JZ8.5 Hz, 2H), 7.78 (d, JZ9.2 Hz, 2H),

11064
P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
amorphous solid (505 mg, 98%). 1H NMR (300 MHz,
CD2Cl2): 10.45 (s, 4H), 7.61 (dd, JZ8.6, 2.7 Hz, 4H), 7.58
(d, JZ2.3 Hz, 4H), 7.45–7.25 (m, 10H), 7.08 (d, 8.6 Hz,
4H), 4.00 (s, 12H). 13C NMR (75 MHz, CD2Cl2): 189.4,
161.5, 143.8, 136.7, 129.9, 129.1, 128.8, 128.5, 127.9,
124.6, 112.4, 84.5, 70.1, 56.3, 55.3. FTIR (neat): 3057,
2943, 2864, 1683, 1604, 1578, 1491, 1463, 1418, 1393,
1282, 1255, 1205, 1180, 1112, 1023 cmK1. HRMS (m/z):
calculated for C50H39O8 [MCH]C: 767.2640, found:
767.2612.
chromatography (SiO2, gradient elution from 30% DCM/
hexanes to 100% DCM then to 100% EtOAc) to give compound
8c as an off white solid (355 mg, 71%). MpZ265–266 8C. 1H
NMR (360 MHz, CDCl3): 7.14–7.12 (m, 12H), 6.83–6.80 (m,
12H), 3.79 (s, 18H). 13C NMR (50 MHz, CDCl3): 158.4, 137.2,
130.1, 113.3, 84.6, 69.4, 55.2, 54.3. FTIR (KBr): 2932, 2836,
2055, 2896, 1742, 1606, 1582, 1502, 1461, 1414, 1300, 1246,
1176, 1114, 1031 cmK1. HRMS (m/z): calculated for C48H42O6
[M]C: 714.2976, found: 715.2971.
5.2.13. 1,4-Bis-(4,40,400-trimethoxytrityl 3,30,300-tricarb-
oxaldehyde)-butadiyne 9c. Bis-trityl 8c (243 mg,
0.339 mmol) and a,a-dichloromethyl methyl ether
(570 mL, 6.1 mmol) were dissolved in DCM (30 mL) and
the resulting solution cooled to 0 8C. TiCl4 (872 mL,
7.7 mmol) was added drop wise via syringe over 2 min.
The reaction was allowed to warm to room temperature and
was stirred an additional 2 hr. The reaction was poured into
aqueous HCl (100 mL, 2.0 M) and stirred for 2 min. The
aqueous suspension was extracted with EtOAc (3!
100 mL). The organic extracts were combined, dried with
MgSO4 and concentrated. The resultant residue was purified
via column chromatography (SiO2, 1:1 DCM–EtOAc) to
give compound 9c as an off white amorphous solid (300 mg,
99%). 1H NMR (300 MHz, CDCl3): 10.39 (s, 6H), 7.55 (d,
JZ2.6 Hz, 6H), 7.49 (dd, JZ8.7, 2.6 Hz, 6H), 6.98 (d,
JZ8.8 Hz, 6H), 3.93 (s, 18H). 13C NMR (75 MHz, CDCl3):
189.4, 161.1, 136.4, 136.0, 128.3, 124.2, 112.0, 83.9, 70.5,
55.8, 54.1. FTIR (neat): 3157, 2944, 2868, 2041, 1682,
1603, 1494, 1462, 1418, 1393, 1281, 1258, 1181, 1113,
1023 cmK1. HRMS (m/z): calculated for C54H43O12 [MC
H]C: 883.2749, found: 883.2716.
5.2.10. 1,4-Diphenyl-1,1,4,4-tetra(2-methoxy-benz-4-yl-
(6-methoxyquinolin-2-ylmethyl)-methylamino)-buta-
diyne 2b. Aldehyde 9b (17 mg, 0.026 mmol), amine 5
(43 mg, 0.21 mmol) were dissolved in DCM (2.0 mL). To
this was added HOAc (10 mL, 0.17 mmol) and NaBH(OAc)3
(45 mg, 0.21 mmol). The reaction was allowed to stir for
10 h. Aqueous NaOH (1 M, 2 mL) was added and the
reaction was stirred for 5 min. The organic layer was
separated and dried with MgSO4. The solvent was removed
and the resultant residue was purified via chromatography
twice (SiO2, 2% TEA/EtOAc then 2% MeOH(NH3)/DCM)
to give compound 2b as an amorphous yellow solid (30 mg,
1
81%). H NMR (300 MHz, CDCl3): 7.86 (d, JZ9.2 Hz,
4H), 7.73 (d, JZ8.6 Hz, 4H), 7.44 (d, JZ2.2 Hz, 4H), 7.40
(d, JZ7.8 Hz, 4H), 7.35–7.15 (m, 14H), 7.06, (dd, JZ8.4,
2.4 Hz, 4H), 6.83 (d, JZ2.7 Hz, 4H), 6.66 (d, JZ8.7 Hz,
4H), 3.83 (s, 12H), 3.75 (s, 8H), 3.69 (s, 12H), 3.56 (s, 8H),
2.13 (s, 12H). 13C NMR (75 MHz, CDCl3): 158.2, 157.2,
156.5, 145.2, 143.3, 136.5, 135.1, 131.0, 130.2, 129.0,
128.4, 128.2, 127.9, 126.7, 126.5, 121.5, 121.2, 109.6,
105.2, 84.8, 69.7, 64.0, 55.4, 55.3, 55.3, 55.2, 42.5. FTIR
(neat): 2940, 2836, 1623, 1601, 1497, 1456, 1378, 1310,
1235, 1161, 1110, 1031 cmK1. HRMS (m/z): calculated for
C98H96N8O2 [MC2H]2C: 756.3670, found: 756.3661.
5.2.14. 1,1,1,4,4,4-Hexa-(2-methoxybenz-4-yl-(6-methoxy-
quinolin-2-yl-methyl) methylamino)-butadiyne 2c. Alde-
hyde 9c (30 mg, 0.033 mmol), amine 5 (81 mg, 0.40 mmol)
were dissolved in DCM (2.0 mL). To this was added HOAc
(23 mL, 0.40 mmol) and NaBH(OAc)3 (85 mg, 0.40 mmol).
The reaction was allowed to stir for 10 h. Aqueous NaOH
(1 M, 2 mL) was added and the reaction was stirred for
5 min. The organic layer was separated and dried with
MgSO4. The solvent was removed and the resultant residue
was purified via chromatography twice on using (SiO2, 2%
TEA/EtOAc then 2% MeOH(NH3)/DCM) to give com-
5.2.11. 4,40,400-Trimethoxytrityl alkyne 7c. Chloride 6c
(2.66 g, 7.14 mmol) was dissolved in argon sparged toluene
(200 mL) and cooled to K78 8C. To this was added ethynyl-
magnesium bromide (30 mL, 0.5 M in ether) via syringe over
10 min. The reaction was allowed to warm to room temperature
and was stirred an additional 3 h. The reaction was quenched
with saturated aqueous NH4Cl and extracted with EtOAC. The
organic layer was dried with MgSO4, filtered, and concentrated
in vacuo. The residue was purified via chromatography (SiO2,
40% DCM/hexanes) to give compound 6c as a white solid
(1.97 g, 76%). MpZ126–128 8C. 1H NMR (300 MHz,
CDCl3): 7.23–7.20 (m, 6H), 6.87–6.84 (m, 6H), 3.82 (s, 9H),
2.71 (s, 1H). 13C NMR (75 MHz, CD2Cl2): 158.3, 137.5, 130.0,
113.2, 90.3, 72.7, 55.2, 53.4. FTIR (neat): 3305, 3154, 2958,
2838, 2253, 1794, 1606, 1583, 1507, 1464, 1381, 1297, 1250,
1178, 1095 cmK1. HRMS (m/z): calculated for C24H22O3
[M]C: 358.1564, found: 358.1560.
1
pound 2c as an amorphous yellow solid (33 mg, 50%). H
NMR (360 MHz, CDCl3): 7.84 (d, JZ9.2 Hz, 6H), 7.67 (d,
JZ8.4 Hz, 6H), 7.47 (d, JZ2.3 Hz, 6H), 7.37 (d, JZ8.8 Hz,
6H), 7.24 (dd, JZ9.2, 2.8 Hz, 6H), 7.10 (dd, JZ8.3, 2.2 Hz,
6H), 6.77 (d, JZ2.5 Hz, 6H), 6.60 (d, JZ8.7 Hz, 6H), 3.81
(s, 18H), 3.69 (s, 12H), 3.63 (s, 18H), 3.52 (s, 12H), 2.09 (s,
18H). 13C NMR (90 MHz, CDCl3): 158.3, 157.2, 156.5,
143.3, 137.1, 135.0, 131.0, 130.3, 128.4, 128.1, 126.5,
121.4, 121.2, 109.7, 105.4, 85.2, 69.7, 64.0, 55.4, 55.3, 55.2,
54.7, 42.5. FTIR (neat): 2935, 2836, 2049, 1681, 1624,
1601, 1499, 1462, 1378, 1310, 1235, 1162, 1110, 1031 cmK1
.
5.2.12. 1,4-Bis-(4,40,400-trimethoxytrityl)-butadiyne 8c.
Alkyne 7c (500 mg, 1.39 mmol) and CuCl (3.45 g,
34.8 mmol) were suspended in DCM (5.0 mL). To this
was added TMEDA (5.85 mL, 38.7 mmol) via syringe over
10 min. The reaction was stirred for 2 days while being
vented to the atmosphere via an 18 gauge needle. The
reaction was quenched with saturated aqueous NH4Cl and
then extracted with DCM. The organic layer dried with
MgSO4 and the solvent removed. The residue was purified via
HRMS (m/z): calculated for C126H128N12O12 [MC2H]2C
1000.4882, found: 1000.4865.
:
Acknowledgements
This work was supported by the National Institutes of
Health (GM 59245).

P. D. Jones, T. E. Glass / Tetrahedron 60 (2004) 11057–11065
11065
Chem. Soc. 1997, 119, 4934–4944. (c) Takeuchi, M.; Shioya,
T.; Swager, T. M. Angew. Chem., Int. Ed. 2001, 40,
3372–3376.
References and notes
1. (a) Chemosensors of Ion and Molecule Recognition; Des-
vergne, J. P., Czarnik, A. W., Eds.; NATO ASI Series C: 492;
Kluwer: Dordrecht, 1997. (b) de Silva, A. P.; Gunaratne,
H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.;
Radermacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97,
1515–1566. (c) Czarnik, A. W. Chem. and Bio. 1995, 2, 423.
(d) Wiskur, S. L.; Ait-Haddou, H.; Lavigne, J. J.; Anslyn, E. V.
Acc. Chem. Res. 2001, 34, 963–972.
6. (a) Marquis, K.; Desvergne, J.-P.; Bouas-Laurent, H. J. Org.
Chem. 1995, 60, 7984–7996. (b) Chen, C.-T.; Huang, W.
J. Am. Chem. Soc. 2002, 124, 6246–6247.
7. (a) Williams, D. H.; Westwell, M. S. Chem. Soc. Rev. 1998,
57. (b) Williams, D. H.; Searle, M.; Mackay, J. P.; Gerhard, U.;
Maplestone, R. A. Proc. Natl. Acad. Sci. U.S.A. 1993, 90,
1172–1178.
2. (a) Glass, T. E. J. Am. Chem. Soc. 2000, 122, 4522–4523. (b)
Raker, J.; Glass, T. E. J. Org. Chem. 2001, 66, 6505–6512.
3. Rebek, J.; Costello, T.; Marshall, L.; Wattley, R.; Gadwood,
R. C.; Onan, K. J. Am. Chem. Soc. 1985, 107, 7481–7487.
4. Shinkai, S.; Ikeda, M.; Sugasaki, A.; Takeuchi, M. Acc. Chem.
Res. 2001, 34, 494–503. Takeuchi, M.; Ikeda, M.; Sugasaki,
A.; Shinkai, S. Acc. Chem. Res. 2001, 34, 865–873.
8. Yang, W.; He, H.; Drueckhammer, D. G. Angew. Chem., Int.
Ed. Engl. 2001, 40, 1714–1718.
9. Raker, J.; Glass, T. E. J. Org. Chem. 2002, 67, 6113–6116.
10. Raker, J.; Glass, T. E. Tetrahedron 2001, 57, 10233–10240.
11. Ikeda, M.; Takeuchi, M.; Sugasaki, A.; Robertson, A.; Imada,
T.; Shinkai, S. Supramol. Chem. 2000, 12, 321.
12. Wyman, J.; Gill, S. J. Binding and Linkage: Functional
Chemistry of Biological Macromolecules; University Science
Books: Mill Valley, CA, 1990.
5. (a) Wipf, P.; Venkatraman, S.; Miller, C. P.; Geib, S. J. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1516–1518. (b) Blanc, S.;
Yakirevitch, P.; Leize, E.; Meyer, M.; Libman, J.; Van
Dorsselaer, A.; Albrecht-Gary, A.-M.; Shanzer, A. J. Am.
13. Connors, K. A. Binding Constants; Wiley: New York, 1987.