Enhancing binding affinity by the cooperativity between host conformation and host-guest interactions

Article abstract of DOI:10.1021/ja203117g

Glutamate-functionalized oligocholate foldamers bound Zn(OAc)₂, guanidine, and even amine compounds with surprisingly high affinities. The conformational change of the hosts during binding was crucial to the enhanced binding affinity. The strongest cooperativity between the conformation and guest-binding occurred when the hosts were unfolded but near the folding-unfolding transition. These results suggest that high binding affinity in molecular recognition may be more easily obtained from large hosts capable of strong cooperative conformational changes instead of those with rigid, preorganized structures.

Full text of DOI:10.1021/ja203117g

COMMUNICATION
pubs.acs.org/JACS
Enhancing Binding Affinity by the Cooperativity between Host
Conformation and HostÀGuest Interactions
Zhenqi Zhong, Xueshu Li, and Yan Zhao*
Department of Chemistry, Iowa State University, Ames, Iowa 50011-3111, United States
S Supporting Information
b
requires phase separation of the polar solvent, folding is most
ABSTRACT: Glutamate-functionalized oligocholate folda-
mers bound Zn(OAc)2, guanidine, and even amine com-
pounds with surprisingly high affinities. The conformational
change of the hosts during binding was crucial to the
enhanced binding affinity. The strongest cooperativity be-
tween the conformation and guest-binding occurred when
the hosts were unfolded but near the foldingÀunfolding
transition. These results suggest that high binding affinity in
molecular recognition may be more easily obtained from
large hosts capable of strong cooperative conformational
changes instead of those with rigid, preorganized structures.
favorable in a solvent mixture with marginal miscibility—e.g., 2:1
hexane/ethyl acetate (EA) with a few percent methanol (MeOH).12
Oligocholates 1À5 were synthesized by methods similar to
those reported previously.9À11 The carboxyl groups were intro-
duced through L-glutamic acids in the foldamer sequence. The
dansyl and naphthyl groups enable us to use fluorescence
resonance energy transfer (FRET) to study the conformation
of the molecules. Folding shortens the distance between the two
fluorophores and allows the excited naphthyl donor to transfer its
energy to the dansyl acceptor. The energy transfer is typically
monitored by the enhanced acceptor emission when the donor is
preferentially excited at 287 nm.9À11
igid supramolecular hosts have been favored traditionally by
R
chemists because of their perceived benefits in binding
affinity. The tradition traces back to Fischer’s lockÀkey theory
and was reinforced by the principle of preorganization articulated
by Cram.1 The idea of preorganization brought great advance-
ment in supramolecular chemistry in recent decades. More
recently, however, there is increasing appreciation that most
biomolecular hosts (e.g., proteins) are folded linear molecules2À5
and not nearly as rigid as preorganized macrocyclic compounds
commonly employed by supramolecular chemists. It is perplex-
ing that, if conformational mobility really represents a disadvan-
tage to high-affinity binding, decades of efforts by chemists
yielded mostly (rigid) synthetic supramolecular hosts that are
no match for their (less rigid) biological counterparts.6 The
answer to this question is significant for a number of important
fields, including enzymatic catalysis and drug development, in
addition to supramolecular chemistry.6 Some researchers, in-
cluding Williams7 and Otto,8 have already started to question
whether chemists, in our efforts to rigidify synthetic hosts, have
wandered away from certain critical elements for success.
We first studied the conformation of 1À3 in MeOH/EA by
fluorescence spectroscopy. This binary mixture represents a
more challenging environment for folding than the ternary
MeOH/(2:1 hexane/EA) mixture.13 According to Figure 2a, 1
was the only compound among the three that could fold in the
binary mixture, evident from its FRET-enhanced dansyl emission
in <15% MeOH. Clearly, the two carboxyl groups were crucial to
the folding. There are two possible ways for the carboxylic acid
groups to help folding. First, they could be involved in dimer-like
hydrogen bonds and stabilize the folded helix directly. Second,
because the folded helix creates a pool of MeOH in its interior
(Figure 1), the polar carboxyl groups are solvated better in the
folded helical conformer than in the unfolded form. The latter
Here we report usual binding behavior of oligocholate folda-
mers. High-affinity binding was obtained with conformationally
mobile hosts, using even relatively weak noncovalent forces,
suggesting that cooperativity or synergism between guest-bind-
ing and the conformational change of a supramolecular host can
be a powerful strategy to enhance hostÀguest interactions.
Oligocholates are amphiphilic foldamers capable of cooperative
conformational changes.9À11 Folding in solution is driven by the
preferential solvation of the cholate hydrophilic faces in a largely
nonpolar solvent mixture (Figure 1). By microphase-separating
some polar solvent molecules and placing them in its internal
nanocavity, the folded oligocholate efficiently satisfies the needs of
both the polar solvent to be located in a polar environment and its
hydrophilic faces to be solvated by polar solvent. Because folding
Received: April 5, 2011
Published: May 16, 2011
r
2011 American Chemical Society
8862
dx.doi.org/10.1021/ja203117g J. Am. Chem. Soc. 2011, 133, 8862–8865
|

Journal of the American Chemical Society
COMMUNICATION
Table 1. Binding Data for Oligocholate 1 at 25 °C
b
entry
guest
% MeOHa
Ka (MÀ1
)
ÀΔG (kcal/mol)
1
2
3
4
5
6
7
8
9
Zn(OAc)2
5
10
15
20
25
15
15
5
(3.1 ( 0.6) Â 105
(8.9 ( 1.0) Â 105
(1.0 ( 0.2) Â 106
(4.1 ( 0.4) Â 105
(2.3 ( 0.4) Â 105
7.5
8.1
8.2
7.7
7.3
Zn(OAc)2
Zn(OAc)2
Zn(OAc)2
Zn(OAc)2
c
c
H2N(CH2)2NH2
H2N(CH2)4NH2
H2N(CH2)6NH2
H2N(CH2)6NH2
À
À
c
c
À
À
Figure 1. Preferential solvation (blue spheres, nonpolar solvent; red
spheres, polar solvent) of a folded oligocholate.
(8.0 ( 1.0) Â 103
(1.5 ( 0.1) Â 104
(1.9 ( 0.2) Â 104
(0.5 ( 0.1) Â 104
5.3
5.7
5.8
5.1
10
15
20
25
5
10 H2N(CH2)6NH2
11 H2N(CH2)6NH2
12 H2N(CH2)6NH2
13 H2N(CH2)8NH2
14 H2N(CH2)8NH2
15 H2N(CH2)8NH2
16 H2N(CH2)8NH2
17 H2N(CH2)8NH2
18 H2N(CH2)10NH2
19 H2N(CH2)12NH2
d
d
À
À
(1.0 ( 0.1) Â 104
(1.0 ( 0.1) Â 104
(2.0 ( 0.2) Â 104
(1.0 ( 0.2) Â 104
5.5
5.5
5.9
5.5
10
15
20
25
15
15
5
d
d
À
À
(1.9 ( 0.3) Â 104
(1.5 ( 0.3) Â 104
(1.9 ( 0.2) Â 105
(3.6 ( 0.4) Â 105
(4.2 ( 0.5) Â 105
(2.7 ( 0.4) Â 105
(1.1 ( 0.2) Â 105
(1.9 ( 0.3) Â 103
5.8
5.7
7.2
7.6
7.7
7.4
6.8
4.5
20
21
22
23
24
6
6
6
6
6
Figure 2. (a) Emission intensity of dansyl at 492 nm for 1 (2), 2 (]),
and 3 (0) in MeOH/EA mixtures. λex = 287 nm; [oligomers] = 2.0 μM.
Corresponding fluorescence spectra are shown in the Supporting Infor-
mation. (b) Binding free energy between 1 and Zn(OAc)2 in MeOH/EA.
10
15
20
25
15
25 6e
a Volume percentage of MeOH in EA. b Association constants deter-
indirectly favors the folded state and was found to assist the
folding of guanidiniumÀcarboxylate-functionalized oligocholates
even when the guanidinium was not involved in the salt bridge.11
Figure 2a shows that complete unfolding of 1 occurred at about
15% MeOH, beyond which all three oligocholates gave similar
emission (i.e., no FRET).14 A large amount of MeOH is known to
diminish the need for the preferential solvation (Figure 1) and
unfolds the oligocholates.9À11 The potential hydrogen bonds
between the carboxyl groups are also weakened by the polar solvent.
Table 1 shows the binding data between 1 and several guests,
determined by fluorescence titration. We studied the binding of
Zn(OAc)2 because a similar dicarboxylated oligocholate was found to
bind the metal in 1:1 stoichiometry.15 In the case of 1, binding
affinities up to 106 MÀ1 were obtained; interestingly, they were higher
in 10À15% MeOH than in 5 or 20% MeOH (Figure 2b).16
The zinc-binding data suggest that the conformation and guest-
binding of the oligocholate are intimately related. The strongest
synergism between the host conformation and the hostÀguest
interactions occurred near the unfoldingÀfolding transition. Initi-
ally, it was unclear to us whether the strong binding in 15% MeOH
was coincidental, but the data suggest that cooperative conforma-
tional change of a host can enhance its binding affinity for the guest.
This is an extremely exciting prospect because, if such synergism can
be rationally engineered, one should be able to achieve high binding
affinity from weak noncovalent forces as long as the binding helps the
host in its conformational change. The strategy is equivalent to
“magnifying” the hostÀguest interactions by the positive coopera-
tivity with the host conformation.
mined by nonlinear least-squares fitting to a 1:1 binding isotherm. λex
=
350 nm. c Binding constant could not be obtained because weak
(5À10%) and random quenching of the dansyl was observed. d Binding
constant could not be obtained because extremely weak quenching
(<5%) was observed. e Host was tetracholate 4.
Figure 3. Binding free energy (a) between 1 and diamines and (b)
between 1 and 6 in MeOH/EA mixtures.
(n = 6À12) caused significant quenching in 15% MeOH.17
Because Ka was quite similar for the longer diamines, we chose
to study hexanediamine and octanediamine in more detail.
The binding constants were surprisingly strong for these flexi-
ble guests (Table 1, entries 8À19). Although lower than those for
Zn(OAc)2, Ka = 104 MÀ1 is quite remarkable for binding driven
by two ammoniumÀcarboxylate ion pairs in up to 20% MeOH.
The binding affinity peaked again at 15% MeOH (Figure 3a).
The effect of MeOH on the ion pairs was unexpected for a polar
solvent. Because of its high polarity and hydrogen-bonding ability,
MeOH normally should weaken the ion pairs continuously with
increasing concentrations.
To test the hypothesis, we turned our attention to several
diamines, H2N(CH2)nNH2, which could only form weak car-
boxylateÀammonium ion pairs with 1 in the polar MeOH/EA
mixtures. Fluorescence titration revealed that the short diamines
(n = 2 or 4) hardly affected the emission of 1, but the longer ones
8863
dx.doi.org/10.1021/ja203117g |J. Am. Chem. Soc. 2011, 133, 8862–8865

Journal of the American Chemical Society
COMMUNICATION
A major difference between zinc- and diamine-binding was
their response to higher amounts of MeOH. In the zinc-binding,
although >15% MeOH lowered the binding affinity and no
binding occurred in neat MeOH, there was only about a 2-fold
decrease in Ka from 20 to 25% MeOH (Table 1, entries 4 and 5).
In the diamine-binding, quenching in 25% MeOH was so weak
(<5%) that the binding constant could not be determined.
These data demonstrate positive cooperativity between the
host conformation and hostÀguest interactions in 1. The syner-
gism is the strongest near the conformational transition and falls
off when the host slips deep into the unfolded region. The results
are reasonable. When the host is folded, strong interactions
already exist between different segments of the host and with
solvents, so guest binding does little to help the intrahost
interactions. When the host is too far in the unfolded region,
binding has to overcome a highly unfavorable conformational
change and is weakened.13 Only when the host is near the
unfoldingÀfolding transition can the host conformation and the
hostÀguest interactions readily help each other.
(Table 1, entries 23 and 24). Not surprisingly, the binding affinity
peaked again in 15% MeOH (Figure 3b).
If strengthened intrahost interactions are the main reason for
enhanced binding affinity, a smaller host with fewer potential
intrahost interactions should bind less strongly when everything
else is equal; i.e., the size of the host should matter greatly in
cooperative binding. The parent oligocholates cannot fold with
fewer than five cholate units.9 Tetracholate 4 has two carboxyl
groups separated by four cholates, the same as in 1. It is unable to
fold in MeOH/EA, evident from the absence of FRET in the
solvent titration (Figure 4S). Indeed, binding between 4 and 6
was much weaker; Ka was at least 2 orders of magnitude lower
than that between 1 and 6 (Table 1, entries 22 and 25).
To further confirm the conformationÀbinding cooperativity, we
synthesized oligocholate 5. WeusedthealkyneÀazide click reaction,26
partly because the synthesis of oligocholates always leaves an azido
group at the chain end.9 Our previous work has demonstrated that
“clicked” oligocholates fold at least as well as the parent compounds.27
The extra carboxylic acid group did not help 5, which was found
to unfold in the MeOH/EA mixture28 but to fold well in <8À10%
MeOH in 2:1 hexane/EA (Figure 5S). It was extremely important
that 5 and 1 had different conformational transitions, or we would
not know whether the strongest cooperativity at the transition point
was general. To our delight, when the binding between 5 and 7 was
studied in the ternary solvent mixtures, the strongest binding once
again occurred at the unfoldingÀfolding transition, this time at ∼8%
MeOH (Figure 6S, Table 1S). In the binary MeOH/EA mixtures, in
which no cooperative conformational change could occur, the
binding not only was weaker but also displayed a slight, monotonous
decrease of Ka with increasing MeOH (Figure 7S). Thus, in the
absence of conformational cooperativity, the normal solvent effect
on the ions pairs dominated.
The above explanation is supported by the discussion of
cooperativity in the literature. “Cooperativity” has a number of
meanings18 and frequently is used when the free energy compo-
nents in a process are not additive. The process is said to have
positive (or negative) cooperativity when the observed free energy
is greater (or smaller) than the sum of the individual contributions.
After studying the “tightening” and “loosening” effects of various
ligands on their biological hosts, Williams and others raised the
interestingpostulation thatthe driving force forguest-bindingdoes
not have to come entirely from direct hostÀguest interactions.7
Instead, a major part may derive from the strengthening of the
existing interactions within the host. Indeed, few biological systems
follow the rule of additivity.19,20 Since conformational changes
frequently accompany guest-binding in protein hosts,21 it is likely
that the strong binding found in protein hosts has major contribu-
tions from the hosts themselves. The notion is supported by the
well-known tight binding between streptavidin and biotion. The
small-molecule ligandraises the protein’s meltingpointby 37°C,22
and a large number of backbone amide protons become resistant
to H/D exchange.23,24 Undoubtedly, the binding has greatly
strengthened the interactions within the host. The enhanced
intrahost interactions very well may be the reason why a small
It is difficult to know exactly where the guests were bound by
the hosts. Because similar solvent effects were found in the
binding of both hydrophilic (zinc acetate) and hydrophobic (1,8-
octanediamine and 1,12-dodecanediamine) guests, the binding
location could not be a determining factor. Our previous studies9
showed that hydrophilic guests prefer to go in the hydrophilic
cavity of the folded helix.29 If displacement of solvent molecules
by the guest was important, one would expect hydrophilic guests
(i.e., ethylenediamine and butylenediamine) to be better guests
and their binding to be stronger in less polar solvents.30
ligand can produce an astounding Ka = 1013.4 MÀ1
.
Our study sheds some important light on the conformationÀ
binding cooperativity. For example, the window for the coopera-
tivity depends on the strength of the hostÀguest binding
interactions. When strong interactions such as ZnÀO complexa-
tion are involved, cooperativity happens over a broad range of
conditions and can tolerate at least 25% MeOH. When weaker
noncovalent forces are involved, the synergism occurs in a much
narrower window, optimal when the host is near the unfol-
dingÀfolding transition. This was probably why the 5% increase
of MeOH from 20 to 25% caused little change to the Zn-binding
but a precipitous drop in Ka for the diamines. Essentially, to
benefit from the conformationÀbinding cooperativity, the guest
needs to fold the unfolded host,25 and such a transition is difficult
when either the host is too far into the unfolded region or the
hostÀguest interactions are too weak.
The selectivity in the binding between 1 and the diamines
warrants some discussion. As far as size is concerned, ethylene-
diamine is more similar to Zn(OAc)2 than octanediamine or
dodecanediamine, and yet it was not bound by 1. Also, why did the
longer diamines (n = 6À12) bind similarly? The answers probably
lie in the nature of the folded oligocholate. Although 1 appears
flexible, much of the cholate is rigid due to ring fusion. The
preferential solvation demands that the hydrophilic faces of the
cholates point inward (Figure 1). These constraints will limit the
movement of the foldamer chain. When strong binding interac-
tions are involved (e.g., ZnÀO complexation), significant strain in
the host can be tolerated during binding, as the major contribution
of the overall binding free energy probably comes from the hostÀ
guest binding interactions. Whenthebindinginteractionsare weak
(e.g., ammoniumÀcarboxylate ion pairing), although significant
binding affinity could be obtained, much of the binding free energy
likely derives from the enhanced intrahost interactions. In the
latter case, the strain introduced by the binding undermines the
intrahost interactions and eliminates the cooperativity. For the
C2ÀC4 dimaines, binding requires the two carboxylic acid groups
The above conclusion was confirmed with diguanidine 6,
which forms stronger salt bridges with carboxylates than the
diamines. With stronger hostÀguest interactions, binding once
again became detectable in 25% MeOH, and changing the
solvent from 20 to 25% MeOH reduced the Ka by only 2.5-fold
8864
dx.doi.org/10.1021/ja203117g |J. Am. Chem. Soc. 2011, 133, 8862–8865

Journal of the American Chemical Society
COMMUNICATION
to be quite close and probably causes significant strain to the
foldamer. For the C6ÀC12 diamines, the distance between the
two amine groups may match reasonably well with the average
distance between the two carboxylic groups in the folded helix.
Similar binding affinity would be expected as long as the main
contributors to the binding (i.e., the ammoniumÀcarboxylate ion
pairs and the enhanced intrahost interactions) are the same.
The principle of preorganization predicts that organizing a host
around its guest through conformational change is detrimental
because the cost of conformational change is assumed to come
from the binding. This prediction is true for supramolecular hosts
incapable of cooperative conformational transition. However, for
large hosts capable of cooperative conformational changes (e.g.,
proteins), our study suggests that the conformation of the host can
be exploited to “magnify” weak binding interactions. Because Ka is
determined by the overall change of free energy during the binding,
one must take into account all the processes that affect free energy.
Guest-induced conformational changes of the host and solvation/
desolvation, whether near or far from the binding site, will affect the
binding affinity. In general, although strong binding interactions are
helpful to establish high binding affinity, they are not necessary. As
long as the guest can trigger a large number of cooperative intrahost
interactions during the binding, high binding affinity will result, as
found for biotinÀstreptavidin.7,31 For this reason, the large size of
biological hosts is not coincidental but critical to their functions.
Another important implication is that chemists and biologists need
to look beyond the targeted binding site when searching for a drug
candidate or strong ligand for a biological host. The best results
likely will come from combined usage of lockÀkey-based techni-
ques (e.g., molecular docking) and those revealing protein dynamics
for the entire structure.
(10) Zhao, Y.; Zhong, Z.; Ryu, E. H. J. Am. Chem. Soc. 2007, 129,
218–225.
(11) Cho, H.; Zhao, Y. J. Am. Chem. Soc. 2010, 132, 9890–9899.
(12) Because the polar solvent is miscible with EA but immiscible
with hexane, a higher percentage of hexane reduces the energetic cost
associated with the phase separation.
(13) Zhao, Y.; Zhong, Z. J. Am. Chem. Soc. 2006, 128, 9988–9989.
(14) Although the emission of dansyl continues to decrease beyond
15% MeOH, the gradual decrease most likely results from the higher overall
polarity of the solvent mixture. Dansyl is known to fluoresce less strongly in
polar solvents. Li, Y. H.; Chan, L. M.; Tyer, L.; Moody, R. T.; Himel, C. M.;
Hercules, D. M. J. Am. Chem. Soc. 1975, 97, 3118–3126.
(15) Zhong, Z.; Zhao, Y. Org. Lett. 2007, 9, 2891–2894.
(16) The binding constants were very high considering that the
carboxylate groups on the oligocholate probably exchanged with the
acetates during binding and thus the strong ZnÀO bonds might not
have contributed directly to the binding enthalpy.
(17) These results indicate that the decrease of the dansyl emission
for the host in the binding studies was a result of specific binding
interactions instead of a generic quenching effect from amines. 1-Hex-
ylamine and 1,6-hexanediol were also tested and found to cause
negligible quenching under the same conditions.
(18) Hunter, C. A.; Anderson, H. L. Angew. Chem. Int. Ed. 2009, 48,
7488–7499.
(19) Dill, K. A. J. Biol. Chem. 1997, 272, 701–704.
(20) Kraut, D. A.; Carroll, K. S.; Herschlag, D. Annu. Rev. Biochem.
2003, 72, 517–571.
(21) Koshland, D. E. Angew. Chem. Int. Ed. Engl. 1995, 33, 2375–2378.
(22) Gonzalez, M.; Bagatolli, L. A.; Echabe, I.; Arrondo, J. L. R.;
Argarana, C. E.; Cantor, C. R.; Fidelio, G. D. J. Biol. Chem. 1997,
272, 11288–11294.
(23) Meskers, S.; Ruysschaert, J.-M.; Goormaghtigh, E. J. Am. Chem.
Soc. 1999, 121, 5115–5122.
(24) Williams, D. H.; Stephens, E.; Zhou, M. J. Mol. Biol. 2003,
329, 389–399.
’ ASSOCIATED CONTENT
(25) Guest-folded host does not have to have the same conforma-
tion as solvent-folded host. Guest-binding was found to increase the
naphthylÀdansyl FRET in 1 only slightly in 15% MeOH. Hence, the
guest-induced folded conformer was less compact than the folded
conformer in 1À2% MeOH/EA. The result is understandable because
guest-induced folding happens in 15% MeOH but the highly compact
folded conformer is stable only at very low levels of MeOH.
(26) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem. Int. Ed. 2002, 41, 2596–2599.
S
Supporting Information. Experimental details; fluores-
b
cence, and binding data. This material is available free of charge
via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
zhaoy@iastate.edu
(27) Pan, X.; Zhao, Y. Org. Lett. 2009, 11, 69–72.
(28) This result suggests that solvation of the carboxyl groups by the
polar solvent in the interior of the folded oligocholate (i.e., the indirect
help to the folding mentioned earlier) is not the main reason for the
strong folding of 1. Otherwise, 5 should fold better than 1.
(29) Itisdifficult to imagine that the long diamines have to stay within the
cavity of the folded helix during binding. Besides, such a binding motif almost
certainly will have a strong dependence on the chain length of the amine guest,
unlike the similar binding affinities found with the C6ÀC12 diamines.
(30) Beyond the unfoldingÀfolding transition, the preferential
solvation shown in Figure 1 is no longer helpful to the host. Before
the transition, the cavity has a concentrated pool of polar solvent. The
unfoldingÀfolding transition, therefore, is where preferential solvation
begins to disappear. From this perspective, release of the solvents should
be less important at the conformational transition than in low-MeOH
mixtures, as the release of phase-separated polar solvent molecules to the
environment is entropically favorable, and this effect will disappear when
the solvent composition in and out of the cavity becomes the same. In
our cholate-based molecular basket, solvent displacement-based binding
was found to weaken with an increasing amount of polar solvent in
the bulk. Zhao, Y.; Ryu, E. H. J. Org. Chem. 2005, 70, 7585–7591.
(31) Weber, P.; Ohlendorf, D.; Wendoloski, J.; Salemme, F. Science
1989, 243, 85–88.
’ ACKNOWLEDGMENT
We thank the U.S. Department of Energy, Office of Basic
Energy Sciences (grant DE-SC0002142), for support.
’ REFERENCES
(1) Cram, D. J. Angew. Chem. Int. Ed. Engl. 1986, 25, 1039–1057.
(2) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173–180.
(3) Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S.
Chem. Rev. 2001, 101, 3893–4012.
(4) Hecht, S.; Huc, I. Foldamers: Structure, Properties, and Applica-
tions; Wiley-VCH: Weinheim, 2007.
(5) Cubberley, M. S.; Iverson, B. L. Curr. Opin. Chem. Biol. 2001,
5, 650–653.
(6) Houk, K. N.; Leach, A. G.; Kim, S. P.; Zhang, X. Y. Angew. Chem.
Int. Ed. 2003, 42, 4872–4897.
(7) Williams, D. H.; Stephens, E.; O’Brien, D. P.; Zhou, M. Angew.
Chem. Int. Ed. 2004, 43, 6596–6616.
(8) (a) Otto, S. Dalton Trans. 2006, 2861–2864. (b) Rodriguez-
Docampo, Z.; Pascu, S. I.; Kubik, S.; Otto, S. J. Am. Chem. Soc. 2006,
128, 11206–11210.
(9) Zhao, Y.; Zhong, Z. J. Am. Chem. Soc. 2005, 127, 17894–17901.
8865
dx.doi.org/10.1021/ja203117g |J. Am. Chem. Soc. 2011, 133, 8862–8865