Building addressable libraries: Site-selective Suzuki reactions on microelectrode arrays

Article abstract of DOI:10.1021/ol900056u

A site-selective Suzuki reaction has been developed for use on microelectrode arrays. The reaction conditions employed are similar to those used to achieve site-selective Heck reactions. The reaction can be run with either an aryliodide attached to the su

Full text of DOI:10.1021/ol900056u

ORGANIC
LETTERS
2009
Vol. 11, No. 6
1273-1276
Building Addressable Libraries:
Site-Selective Suzuki Reactions on
Microelectrode Arrays
Libo Hu,† Karl Maurer,‡ and Kevin D. Moeller*,†
Department of Chemistry, Washington UniVersity, St. Louis, Missouri 63130 and, and
CombiMatrix Corporation, 6500 Harbor Heights Parkway, Suite 301,
Mukilteo, Washington 98275
moeller@wuchem.wustl.edu
Received January 12, 2009
ABSTRACT
A site-selective Suzuki reaction has been developed for use on microelectrode arrays. The reaction conditions employed are similar to those
used to achieve site-selective Heck reactions. The reaction can be run with either an aryliodide attached to the surface of the array and an
arylboronic acid in solution or with an arylboronic acid attached to the surface of the array and an arylbromide in solution. Both allyl acetate
and air are effective confining agents for the reaction. The reactions are compatible with arrays containing either 1024 microelectrodes cm-2
or 12 544 microelectrodes cm-2
.
Addressable microelectrode arrays1,2 are intriguing tools for
probing interactions between small molecules and biological
receptors3,4 because they enable monitoring of the interac-
tions in “real-time”.5 For this reason, we have been develop-
ing synthetic methods for site-selectively building molecules
proximal to the electrodes in a microelectrode array.6,7 The
long-term goal of the work is to use the arrays as a platform
for small molecular libraries containing engineered, semirigid
peptidomimetics that are useful for probing the conforma-
tional requirements of protein receptors.
Triaryl peptidomimetics that mimic helical regions of
proteins8,9 are of interest in this context. Helices on protein
surfaces often play key roles in the binding of proteins to
(3) For the use of small molecule arrays for probing biological receptors,
see: (a) Lipshutz, R. J.; Fodor, S. P. A.; Gingeras, T. R.; Lockhart, D. J.
Nat. Genet. 1999, 21, 20. (b) Pirrung, M. C. Chem. ReV. 1997, 97, 473.
For more recent examples, see: (c) Webb, M. S.; Miller, A. L.; Johnson,
B. H.; Fofanov, Y.; Li, T.; Wood, T. G.; Thompson, E. B. J. Steroid
Biochem. Mol. Biol. 2003, 85, 183. (d) Shih, S.-R.; Wang, Y.-W.; Chen,
G.-W.; Chang, L.-Y.; Lin, T.-Y.; Tseng, M.-C.; Chiang, C.; Tsao, K.-C.;
Huang, C. G.; Shio, M.-R.; Tai, J.-H.; Wang, S.-H.; Kuo, T.-L.; Liu, W.-T.
† Washington University.
‡ CombiMatrix Corporation.
(1) For a description of the microelectrode array chips used here, see:
Dill, K.; Montgomery, D. D.; Wang, W.; Tsai, J. C. Anal. Chim. Acta 2001,
J. Virol. Methods 2003, 111, 55
.
444, 69
.
(4) For reviews, see: (a) Duffner, J. L.; Clemons, P. A.; Koehler, A. N.
(2) For alternative approaches using electrochemical microarrays, see:
(a) Sullivan, M. G.; Utomo, H.; Fagan, P. J.; Ward, M. D. Anal. Chem.
1999, 71, 4369. (b) Zhang, S.; Zhao, H.; John, R. Anal. Chim. Acta 2000,
421, 175. (c) Hintsche, R.; Albers, J.; Bernt, H.; Eder, A. Electroanalysis
Curr. Opin. Chem. Biol. 2007, 11, 74. (b) Taipa, M. A. Comb. Chem. High
Throughput Screening 2008, 11, 325
(5) (a) Tesfu, E.; Roth, K.; Maurer, K.; Moeller, K. D. Org. Lett. 2006,
8, 709. (b) Stuart, M.; Maurer, K.; Moeller, K. D. Bioconjugate Chem.
.
2000, 12, 660
.
2008, 19, 1514.
10.1021/ol900056u CCC: $40.75
Published on Web 02/24/2009
2009 American Chemical Society

other proteins, DNA, and RNA. Hence, they represent
attractive targets for developing new methods for treating a
variety of diseases. Triaryl ring systems can mimic sections
of these helices by orientating peptide side-chains in either
the i, i+3, and i+7 or i, i+4, and i+7 positions. They have
the potential to be particularly useful in this regard because
they represent typical drug-like scaffolds. With this in mind,
we became interested in developing the synthetic methodol-
ogy needed to build microelectrode array-supported libraries
of triaryl peptidomimetics.
The study of site-selective Suzuki reactions began with
the placement of a 4-iodobenzoate substrate (1) onto the
surface of an array having 1024 microelectrodes cm-2
(Scheme 1). The substrate was placed proximal to each of
Scheme 1
The key to constructing triaryl ring skeletons is the Suzuki
reaction.8,10 The Suzuki reaction allows for a “building-
block” approach to the molecules using preconstructed,
individual aryl rings that are then coupled to each other. But
can such an approach be employed site-selectively on a
microelectrode array? Microelectrode array reactions involve
a delicate balance between the electrochemical generation
of a catalyst, the rate of the desired reaction, and the
destruction of the catalyst generated in the solution above
the array by a “confining” agent. In this balance, the
confining agent must be reactive enough to prevent migration
of the catalyst to nonselected sites in the array. This is a
challenging task since the arrays can contain up to 12 544
microelectrodes cm-2. On the other hand, the confining agent
must not be too reactive or it will prevent the desired reaction
from occurring at the selected electrodes. Hence, the critical
step in the development of any new reaction on a micro-
electrode array is identifying an effective confining agent.
For the Suzuki reaction, the need for a confining agent
immediately raised questions about the generality of the
strategy used to confine the earlier Pd(0)-catalyzed reactions.7
Does changing the reaction catalyzed by a reagent on a
microelectrode array require the identification of a new
confining agent, or can the parameters of the electrolysis be
adjusted to account for differences in the reactions so that
the same confining agents can be employed? The answer to
this question is essential for planning new reactions and
synthetic strategies on the arrays. We report here that site-
selective Suzuki reactions can be accomplished using the
confining strategy developed earlier for the Heck reaction if
the rate of Pd(0) generation at the microelectrodes is reduced
and the concentration of the confining agent increased to
account for a faster reaction.
the electrodes in the array using the electrogenerated base
catalyzed procedure previously developed for coupling
activated esters to the agarose polymer coating the array.6,7
The Suzuki reaction was then initiated using nearly the same
conditions utilized to accomplish site-selective Heck
reactions.7a Accordingly, the entire surface of the array was
treated with an acetonitrile/dimethylformamide/water solution
containing pyrene-1-boronic acid 2, palladium acetate, triph-
enylphosphine, and tetrabutylammonium bromide. The Pd(II)
reagent was then reduced to the required Pd(0)-catalyst at
selected locations on the array by using the microelectrodes
as cathodes. The potential at the selected electrodes was set
at -2.4 V relative to a remote Pt-wire anode for 0.5 s and
then the electrodes turned off for 0.1 s. Nonselected
electrodes were left off for the entire reaction. The only
change in the reaction relative to the site-selective Heck
reaction was the use of allyl acetate as the confining agent
instead of the allyl methyl carbonate employed previously.
The confining agent is added to the solution above the array
in order to oxidize any Pd(0) that begins to migrate away
from the microelectrodes selected as cathodes. This prevents
unwanted reactions from occurring proximal to electrodes
that were not selected. Both allyl acetate and allyl methyl
(6) For Pd(II) reactions, see: (a) Tesfu, E.; Maurer, K.; Ragsdale, S. R.;
Moeller, K. D. J. Am. Chem. Soc. 2004, 126, 6212–6213. (b) Tesfu, E.;
Maurer, K.; McShae, A.; Moeller, K. D. J. Am. Chem. Soc. 2006, 128, 70.
For examples using electrogenerated base, see: Maurer, K.; McShea, A.;
Strathmann, M.; Dill, K. J. Comb. Chem. 2005, 7, 637. For the site-selective
generation of acid, see: Kesselring, D.; Maurer, K.; Moeller, K. D. Org.
Lett. 2008, 10, 2501. For the use site-selective use of ceric ammonium
nitrate, see: Kesselring, D.; Maurer, K.; Moeller, K. D. J. Am. Chem. Soc.
2008, 130, 11290
.
(7) For a Pd(0) reactions see: (a) Tian, J.; Maurer, K.; Tesfu, E.; Moeller,
K. D. J. Am. Chem. Soc. 2005, 127, 1392. (b) Tang, F.; Chen, C.; Moeller,
K. D. Synthesis 2007, 3411. (c) Chen, C.; Lu, P.; Walker, A.; Maurer, K.;
Moeller, K. D. Electrochem. Commun. 2008, 10, 973. (d) Tian, J.; Maurer,
Scheme 2
K.; Moeller, K. D. Tetrahedron Lett. 2008, 49, 5664
.
(8) For the use of triaryl systems as helix mimics, see: Yin, H.; Lee,
G.; Seedy, K. A.; Kutzki, O.; Park, H. S.; Orner, B. P.; Ernst, J. T.; Wang,
H.-G.; Sebti, S. M.; Hamilton, A. D. J. Am. Chem. Soc. 2005, 127, 10192
(9) For the phenyl dipyridyl systems, see: Che, Y.; Brooks, B. R.;
Marshall, G. R. Biopolymers 2007, 86 (4), 288
(10) For 2: Unpublished results with Bourne, G. and Marshall, G. R.
.
.
1274
Org. Lett., Vol. 11, No. 6, 2009

carbonate work as confining agents by converting the Pd(0)
catalyst into a π-allyl-Pd(II) species (Scheme 2). The switch
to allyl acetate was made to avoid the generation of
methoxide during this process. In the Heck reaction, it was
thought that the methoxide generated cleaved the ester
connecting the product to the agarose polymer.7a This
decreased the amount of product on the surface of the array
with time. The switch to allyl acetate was made to eliminate
this possibility.
The use of these conditions with a checkerboard pattern
of microelectrodes employed as cathodes led to a moderate
degree of success (Figure 1a). When the array used to
Figure 2. Fluorescence image of an experiment that ran the VB12
esterification reaction in box pattern and the Suzuki reaction on all
of the electrodes in the array. The spots at the top of the image are
part of a second box pattern.
completion of the reaction, the microelectrode array was
examined with a fluorescence microscope giving rise to the
image shown in Figure 2. Only the box of electrodes employed
in the initial esterification reaction showed the presence of
pyrene. There was no background reaction between the pyrene
boronic acid and the agarose polymer.
With the reactions working well on arrays having 1024
microelectrodes cm-2 (a 1 K-array), attention was turned to
the use of more dense arrays having 12 544 microelectrodes
cm-2 (a 12 K-array). This work was done because it is the
more dense arrays that are used in signaling studies.5 Initially,
the experiment on the 12 K-array was attempted using
reaction conditions that were identical to those used on the
1 K-array. In this case, the initial substrate placement reaction
was performed at each microelectrode in the array. The
Suzuki reaction was then initiated using a checkerboard
inside of a box pattern of microelectrodes. While the
checkerboard pattern could be seen in places on the array,
confinement was lost leading to reactions at electrodes that
were not selected (Figure 3a). The pattern was difficult to
Figure 1. Fluorescence image of the site-selective Suzuki reaction
(a) checkerboard pattern run using -2.4 V (b) checkerboard pattern
run using -1.7 V.
conduct the Suzuki reaction was imaged using a fluorescence
microscope, a checkerboard pattern for the pyrene product
could be seen. However, significant reaction also occurred
at electrodes that were not selected. The checkerboard pattern
was only visible because the pyrene-derived fluorescence
associated with the selected electrodes was brighter than that
associated with the nonselected electrodes. Clearly, confine-
ment on the array was being lost and migration of the Pd(0)
generated at the selected electrodes was not being completely
prevented. The loss of confinement for the Suzuki reaction
relative to the Heck reaction run earlier suggested that either
the Suzuki reaction is faster than the Heck reaction, the allyl
acetate is less effective as a confining agent than the allyl
methylcarbonate employed in the Heck reaction, or both.
To stop the migration of Pd(0) to microelectrodes not selected
for the Suzuki reaction, the rate of Pd(0) generation at the
selected electrodes was decreased. This was accomplished by
reducing the potential at the electrodes from -2.4 V to -1.7
V relative to the remote Pt counter electrode. Once again, the
selected electrodes were cycled on for 0.5 s and off for 0.1 s.
The change resulted in far superior confinement (Figure 1b).
In this case, no fluorescence was observed at electrodes not
selected as cathodes for the generation of Pd(0).
Figure 3. Fluorescence image of a site-selective Suzuki reaction
on a 12 K-array: (a) checkerboard in a box pattern run with the
conditions developed for the 1 K-array; (b) same experiment run
with double the confining reagent.
To demonstrate that the fluorescence observed in Figure 2
was a result of the Suzuki reaction and not a reaction between
the pyrenyl boronic acid and the agarose polymer coating the
array, a control experiment was performed. In this study, the
initial base catalyzed esterification reaction was performed using
a boxes of eight electrodes. The Suzuki reaction was performed
using all of the electrodes in the array as cathodes. After
observe. For example, compare the array illustrated in Figure
3a with the array in Figure 3b that clearly shows the
checkerboard in a box pattern. The image shown in Fig-
ure 3b was obtained by doubling the concentration of allyl
Org. Lett., Vol. 11, No. 6, 2009
1275

for Pd(0), will have no effect on this reactive intermediate.
Instead confinement of the aryl-Pd(II) species depends solely
on the relative rates of diffusion away from the selected cathode
and capture of the aryl-Pd(II) intermediate by the boronic acid
on the surface of the array. If the rate of the Suzuki reaction is
significantly faster than diffusion away from the electrode, then
the reaction will be confined. If it is not, then the reaction will
lose confinement. In practice, the reaction was performed in
the same manner as the earlier 1 K-array reaction. A high level
of confinement to the selected microelectrodes was obtained
(Figure 5) indicating that the Suzuki reaction is fast enough to
Figure 4. Fluorescence image of an air confined Suzuki reaction
run on a 1 K-array using a potential at the working electrode of
-1.4 V relative to the counter electrode.
acetate confining agent used for the reaction from 0.54 M to
1.08 M. Clearly, site-selective Suzuki reactions can be
accomplished on a 12 K-microelectrode array.
Next, a Suzuki reaction was run using air as the confining
agent. No allyl acetate was added to the reaction. In this
case, the current used for the reaction needed to be reduced
further in order to maintain confinement to a preselected
electrode. This was accomplished by decreasing the potential
at the working microelectrode to -1.4 V relative to the
counter electrode (Figure 4). In this experiment, a single
microelectrode in a 1 K-array was used as a cathode to
generate the Pd(0) catalyst. The need to reduce the rate at
which the Pd(0) was generated in order to maintain confine-
ment suggests that oxygen is not as an effective scavenger
for Pd(0) as is the allyl acetate used in the previous
experiments.
Figure 5. Fluorescence image of the inverse Suzuki reaction.
avoid migration of the aryl-Pd(II) intermediate away from its
point of origin. Unlike the earlier Suzuki reaction, the “inverse-
Suzuki” worked well even when the Pd(0) reagent needed is
generated at a fast rate (electrode potential of -2.4 V vs the
counter electrode). This observation appears to be a result of
the high concentration of bromopyrene used in the reaction.
The bromopyrene also reacts with the Pd(0) generated to prevent
its migration to neighboring electrodes.
In conclusion, the reaction conditions developed for achieving
site-selective Heck reactions on a microelectrode array have
been successfully extended to the Suzuki reaction. Both allyl
acetate and oxygen can be used as confining agents for the faster
Suzuki reaction if the rate of reagent generation is limited and
the concentration on the confining agent raised. The reactions
work well with either an aryliodide or an arylboronic acid
attached to the surface of the microelectrode array and the other
coupling partner in solution. Finally, site-selective Suzuki
reactions can be accomplished on either 1 K- or 12 K-arrays.
Finally, an “inverse-Suzuki” reaction was done on a 1
K-microelectrode array. In this experiment, a boronic acid
coupling partner was placed on the array and a bromopyrene
used as the aromatic ring in solution (Scheme 3). The challenge
Scheme 3
Acknowledgment. We thank the National Science Foun-
dation (CHE-0613077) for their generous support of our
work. We also gratefully acknowledge the Washington
University High Resolution NMR facility, partially supported
by NIH grants RR02004, RR05018, and RR07155, and the
Washington University Mass Spectrometry Resource Center,
partially supported by NIHRR00954, for their assistance.
Supporting Information Available: Sample experimental
procedures are included for both placing substrates onto the
microelectrode arrays and the site-selective Suzuki reaction
along with characterization data for compound 1. This
material is available free of charge via the Internet at
http://pubs.acs.org.
of running the reaction in this direction is that the generation
of Pd(0) at the selected microelectrodes is followed by inser-
tion of the catalyst into a solution phase aryl bromide bond.
This results in a reactive aryl-Pd(II) species that is not tied to
the surface of the array. The confining agent, which is an oxidant
OL900056U
1276
Org. Lett., Vol. 11, No. 6, 2009