Design of chemical shift-switching 19F magnetic resonance imaging probe for specific detection of human monoamine oxidase A

Article abstract of DOI:10.1021/ja2057506

Monoamine oxidase (MAO) A is a flavoenzyme that catalyzes the oxidation of biologically important monoamines and is thought to be associated with psychiatric disorders. Here, we report a strategy for rationally designing a ¹⁹F magnetic resonanc

Full text of DOI:10.1021/ja2057506

COMMUNICATION
pubs.acs.org/JACS
Design of Chemical Shift-Switching 19F Magnetic Resonance Imaging
Probe for Specific Detection of Human Monoamine Oxidase A
Koya Yamaguchi,† Ryosuke Ueki,† Hiroshi Nonaka,† Fuminori Sugihara,‡ Tetsuya Matsuda,‡ and
Shinsuke Sando*,†,§
†INAMORI Frontier Research Center, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
‡Department of Systems Science, Graduate School of Informatics, Kyoto University, 36-1 Yoshida-Honmachi, Sakyo-ku,
Kyoto 606-8501, Japan
§PRESTO, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan
S Supporting Information
b
Scheme 1. Mechanism of Oxidative Deamination by MAO
ABSTRACT: Monoamine oxidase (MAO) A is a flavoen-
zyme that catalyzes the oxidation of biologically important
monoamines and is thought to be associated with psychiatric
disorders. Here, we report a strategy for rationally designing a
19F magnetic resonance imaging probe for the specific detec-
because of the high sensitivity of 19F (0.83 relative to H) and
extremely low background signals in biological samples.6À9
19F-labeled compounds, such as fluorinated serotonin7a or dopa-
mine,7b have potential for use in an 19F-based MAO assay. To our
knowledge, however, none of these has been shown to have
MAO-A/B selectivity or to induce sufficient 19Fsignalchangesfor19F
magnetic resonance imaging (MRI). Under these circumstances, we
were prompted to develop an hMAO-A-specific MRI probe. Here, we
report a strategy for rationally designing a 19F MRI probe targeting
hMAO-A. Our probe achieved high specificity to hMAO-A and
realized the 19F chemical shift-based imaging of hMAO-A activity.
When designing an hMAO-A-specific 19F MRI probe, there
are two important issues to be solved. The first issue is how to
incorporate hMAO-A specificity into the probe. hMAO-A and -B
have about 70% identity,1 and discriminating between these
two isoforms is a challenge. We focused on hMAO inhibitors.
Because hMAO inhibitors are used as potent pharmaceutical
agents to treat psychiatric or neurodegenerative disorders, selec-
tive inhibitors have been investigated vigorously.1b Clorgyline
(Figure 1a, left) is a known hMAO-A-selective inhibitor.1a
Interestingly, structurally related pargyline (Figure 1a, second
from left) is an inhibitor but is specific to hMAO-B.1a On the
basis of this comparison, we hypothesized that ortho/para-
substituted phenol (Figure 1a, third from left) is a key scaffold
to achieve hMAO-A-selective binding.
1
tion of human MAO-A (hMAO-A) activity. Our designed 19F
probe was oxidized expeditiously by hMAO-A to produce
2-fluoro-4-nitrophenol via a spontaneous β-elimination me-
chanism. Concomitant with the structural change of the probe
to the product, the 19F chemical shift changed by 4.2 ppm,
which was enough to visualize the probe and enzymatic
product separately. Importantly, our probe achieved excellent
discrimination of hMAO-A from its isoform hMAO-B.
onoamine oxidase (MAO) is a flavoenzyme that catalyzes
M
the oxidation of monoamines to the corresponding imi-
nium intermediates, which are subsequently hydrolyzed to
aldehydes (Scheme 1).1 In humans, MAO exists as two iso-
forms: hMAO-A and hMAO-B. These isoforms play different
physiological roles.
hMAO-A oxidizes biologically important monoamine neuro-
transmitters, including serotonin and epinephrine, to maintain
homeostasis. hMAO-A activity is thought to be associated with
psychiatric disorders.1 Current research suggests that brain
hMAO-A increases in major depressive disorder (MDD), and
elevated hMAO-A levels could be the primary cause of the
monoamine-lowering process in MDD.2 In fact, hMAO-A-selec-
tive inhibitors are prescribed as clinical antidepressant agents. By
contrast, hMAO-B-selective inhibitors are devoid of such anti-
depressant activity.2 In this sense, hMAO-A activity is one of the
primary targets for psychiatric disorder analysis, and thus there is
increasing focus on chemical probes that can selectively detect
hMAO-A activity in a complex biological system.
The second issue is how to convert the o/p-substituted
phenol, a hypothesized hMAO-A-selective binder, to the 19F
MRI probe. For this purpose, we designed compounds 1À3 as
potent 19F MRI probes with hMAO-A specificity (Figure 1a,
right). These probes are designed to work as a 19F chemical shift-
switching probe via the following mechanism (see Figure 1b).
After enzymatic oxidation of an amino group to aldehyde by
MAO, a propionaldehyde moiety is released spontaneously by
β-elimination in water to afford o/p-19F/NO2-phenol. Because of
the electron-withdrawing property of the NO2 substituent, the
To date, several optical hMAO indicators have been reported,3À5
and some of these have realized the sensitive detection of hMAO
activity in cells. However, these optical modality-based approaches
typically suffer from difficulties when applied to ex vivo or in vivo
applications because of the limited transparency of excitation and
emission light through the opaque body.
In this context, magnetic resonance (MR) is an appropriate
modality because MR signals can be detected even in deeper
sites. In particular, a 19F MR-based probe is most attractive
Received: June 21, 2011
Published: August 18, 2011
r
2011 American Chemical Society
14208
dx.doi.org/10.1021/ja2057506 J. Am. Chem. Soc. 2011, 133, 14208–14211
|

Journal of the American Chemical Society
COMMUNICATION
Figure 1. (a) Design of hMAO-A-specific 19F MRI probes 1À3. (b)
Proposed mechanism for the spontaneous production of o/p-substituted
phenol from 19F MRI probes 1À3 upon oxidation by hMAO-A.
pKa of the produced fluoronitrophenol is <7 (e.g., pKa = 6.03 for
2-fluoro-4-nitrophenol9c), and the product should be present
mainly as the phenolate form under physiological conditions.
Thus, the net conversion from probes 1À3 to the o/p-19F/NO2-
substituted phenolate ion should induce a sufficient chemical
shift change of 19F on the phenyl ring, as suggested by reports
from Mason’s group of a 19F MRI reporter protein using an
o/p-19F/NO2-phenol-based substrate.9
We first evaluated the reactivity of probes 1À3 to hMAO-A.
Typically, the enzymatic reaction of probe (100 μM) by hMAO
(12 units/mL) was performed in HEPES buffer (100 mM, pH
7.4) at 37 °C. After incubation with hMAO-A, intact probes were
quantified by HPLC-UV analysis. During 1 h incubation, 65.8 (
0.7, 10.5 ( 0.9, and 46.9 ( 1.1% of probes 1, 2, and 3 were
consumed, respectively, suggesting that all three probes can be a
substrate of hMAO-A and probe 1, having a nitro substituent at
the para position, is the best substrate among these probes
(Figure 2a). This p-NO2 selectivity is reasonable, based on
crystallographic data of the clorgylineÀhMAO-A complex,
which shows a relatively wide space to accommodate the larger
nitro group at the para position (Figure 2b).10
Figure 2. (a) Reactivity of probes 1À3 (100 μM) with hMAO-A (12
units/mL) in HEPES buffer (100 mM, pH 7.4) at 37 °C. The fraction of
consumed probe was determined by HPLC-UV analysis of each enzy-
matic reaction solution. Error bars indicate the standard deviation of
three independent experiments. (b) Active pocket of hMAO-A com-
plexed with the inhibitor clorgyline and cofactor flavin adenine dinu-
cleotide (FAD) (PDB: 2BXR). The chemical structure of clorgyline is
shown to the left. The O, N, and Cl of clorgyline are depicted in red, blue,
and green, respectively; FAD is shown in yellow. (c) HPLC-UV analyses
(312 and 400 nm) of probe 1 (100 μM) after 1 h incubation: (top left)
without hMAO-A, (top right) with hMAO-A (12 units/mL), (bottom
left) with hMAO-B (12 units/mL), and (bottom right) with MAO-A
(12 units/mL) and clorgyline (10 μM).
We focused further on the enzymatic reactivity of probe 1.
HPLC-UV analyses were performed at the λmax of probe 1
(312 nm) and λmax of the expected product 2-fluoro-4-nitrophe-
nol (400 nm) (Figure 2c). After 1 h incubation with hMAO-A,
the HPLC peak of probe 1 (15.8 min) was clearly reduced (top
left vs top right at 312 nm). Monitoring at the λmax of the
enzymatic product (400 nm) showed concomitant appearance of
a new peak at 10.6 min (Figure 2c, top right). The new peak was
confirmed as the predicted 2-fluoro-4-nitrophenol by compar-
ison with an authentic sample. Calculated from the peak in-
tensity, the molar amount of 2-fluoro-4-nitrophenol produced
corresponded to 66.0 ( 0.3% of probe 1, indicating that about
100% of the consumed starting probe 1 was oxidized and
converted expeditiously to 2-fluoro-4-nitrophenol via sponta-
neous β-elimination. These results validate the designed hMAO-
A-induced deamination and β-elimination mechanism.
MAO-A inhibitor clorgyline (10 μM), production of 2-fluoro-4-
nitrophenol was suppressed completely (Figure 2c, bottom
right). The designed probe 1 worked as a substrate of hMAO-
A with excellent specificity.
The enzymatic production of 2-fluoro-4-nitrophenol from
probe 1 was also monitored by measuring the absorption. Upon
reaction with hMAO-A (12 units/mL), the absorption spectra of
probe 1 (100 μM) changed in an incubation time-dependent
manner with a single isosbestic point at 345 nm (Figure 3a). The
product 2-fluoro-4-nitrophenol (pKa = 6.039c) presented mainly
in the phenolate ion form at physiological pH, so the product had
an absorption maximum at a relatively long wavelength of
400 nm (ε400 nm = 17 200 MÀ1 cmÀ1 in HEPES (100 mM, pH
7.4)). By contrast, the o-alkylated starting probe 1 had a low
extinction coefficient at 400 nm. Therefore, the enzymatic
conversion of 1 to product could be monitored easily by the
absorption change at 400 nm. As shown in Figure 3b, a clear
increase in absorption was observed for the probe 1 solution in
the presence of hMAO-A (red circles). The presence of hMAO-B
(green circles) or hMAO-A with the inhibitor clorgyline (blue
circles) caused almost no absorption change. This specific
Very importantly, probe 1 showed high specificity for hMAO-
A over hMAO-B. In contrast to the clear conversion by hMAO-A,
only a faint amount (<1%) of the product was observed after
incubation with hMAO-B (Figure 2c, bottom left). When the
reaction with hMAO-A was performed in the presence of the
14209
dx.doi.org/10.1021/ja2057506 |J. Am. Chem. Soc. 2011, 133, 14208–14211

Journal of the American Chemical Society
COMMUNICATION
Figure 3. (a) Absorption spectra of probe 1 (100 μM) incubated with
hMAO-A (12 units/mL) in HEPES buffer (pH 7.4) at 37 °C for 0, 20,
40, 60, and 90 min (from bottom to top at 400 nm). (b) Time course of
the absorption change at 400 nm of probe 1 (100 μM) solution in the
presence of hMAO-A (12 units/mL) (red), hMAO-A with the inhibitor
clorgyline (10 μM) (blue), or hMAO-B (green). (c) Image of probe 1
solution (500 μM) incubated in the presence of hMAO-A (60 units/
mL), hMAO-A with the inhibitor clorgyline (50 μM), and hMAO-B
(from bottom to top) for 0, 20, 40, 60, and 90 min (from left to right).
reactivity of probe 1 to hMAO-A is consistent with the results
obtained by HPLC-UV analysis (Figure 2).
Using the MichaelisÀMenten equation, the kinetic para-
meters for enzymatic reaction of hMAO-A with probe 1 were
determined as Km = 37.2 ( 1.7 μM and Vmax = 24.1 ( 0.8 nmol
minÀ1 mg of microsomal proteinÀ1. The obtained Km value is
lower than that of natural monoamine substrates (high μMÀmM
range, e.g., Km = 137 μM for serotonin1b), showing that probe 1
can be a practical substrate for hMAO-A under physiological
conditions. By contrast, the reactivity of probe 1 with hMAO-B
was too small to determine these parameters.
Figure 4. (a) 19F NMR spectra of probe 1 and 2-fluoro-4-nitrophenol
1
(100 μM) in HEPES buffer (pH 7.4). (b) H and 19F chemical shift-
selective imaging (11.7 T) of probe 1 and 2-fluoro-4-nitrophenol (5 mM
each) in HEPES buffer (pH 7.4). (c) Time course (0À90 min) of
19F NMR spectra of probe 1 (100 μM) upon reaction with hMAO-A
(12 units/mL). (d) 1H and 19F chemical shift-selective imaging (11.7 T)
of hMAO-A activity (240 units/mL) using probe 1 (5 mM) in the
presence or absence of the inhibitor clorgyline (250 μM).
Because of the appreciable absorption enhancement at
400 nm, hMAO-A activity could be detected easily by the
unaided eye. Figure 3c shows a photograph of probe 1 solutions
in the presence of hMAO-A (bottom line), hMAO-A with
clorgyline (middle line), and hMAO-B (top line). In the presence
of hMAO-A, probe 1 solution turned yellow in an incubation
time-dependent manner (from 0 to 90 min, from left to right). In
marked contrast, the other solutions remained colorless. This
indicates that probe 1 can also be used as a chromogenic probe
for the specific sensing of hMAO-A activity. Although such
chromogenic sensing is not our present concern, we note that
this probe allows a mix-and-read sensing without the need for
additional reagents, enzymes, or steps for monitoring. The
hMAO-A/B selectivity, reactivity, and simplicity are superior to
those reported for other chromogenic probes.5
Having an hMAO-A-specific substrate in hand, we moved on
to the detection of hMAO-A activity using the 19F MR modality.
19F NMR analyses of probe 1 and product 2-fluoro-4-nitrophe-
nol gave 19F signals at À133.3 and À137.5 ppm, respectively
(Figure 4a). The observed chemical shift difference between the
probe and product was 4.2 ppm, which was enough to visualize each
compound selectively using 19F MRI. Figure 4b presents the
phantom images (11.7 T) of these compounds in tubes. 1H MRI
visualized both tubes because of the presence of a large amount of
H2O (Figure 4b, left). By contrast, probe 1 and product 2-fluoro-4-
nitrophenol were visualized separately by 19F chemical shift-selective
imaging (Figure 4b, middle and right, obtained by probe 1 and 2-
fluoro-4-nitrophenol selective pulse frequencies, respectively).11
Finally, we applied probe 1 in 19F MR analysis of hMAO
activity. Probe 1 was incubated with hMAO-A, and the mixture
was subjected to 19F NMR analysis at each time point (0À
90 min). At the starting point (0 min), a single 19F peak was
observed at À133.3 ppm (Figure 4c, bottom). As the reaction
proceeded, a new signal appeared at À137.5 ppm in an incuba-
tion time-dependent manner (Figure 4c, from bottom to top,
0À90 min), which was assigned as the produced 2-fluoro-4-
nitrophenol by comparison with an authentic sample. The time-
course profile of 19F NMR (79% conversion of probe 1 to
product after the 90 min reaction, determined by comparison of
the 19F peak integrals) was similar to that observed by absorption
analysis (Figure 3b) or HPLC-UV analysis (Figure S1, 78%
conversion after 90 min, determined by HPLC), supporting that
the 19F NMR-based approach produces quantitative results.
Again importantly, the new 19F NMR peak at À137.5 ppm was
14210
dx.doi.org/10.1021/ja2057506 |J. Am. Chem. Soc. 2011, 133, 14208–14211

Journal of the American Chemical Society
COMMUNICATION
hMAO-A dependent. No signal was observed around À137.5
ppm after 60 min incubation with hMAO-B or hMAO-A in the
presence of clorgyline (Figure S2). In addition to the hMAO-A
specificity, it should be also noticed that a new NMR peak
appeared in a clear probe-to-product one-to-one conversion
manner without any undesirable side peaks. This is a big benefit
for chemical shift-selective imaging, where sometimes such side
peaks result in nonnegligible background or false-positive con-
trast. The designed probe 1 realized the direct imaging of hMAO-A
activity (Figure 4d). 19F chemical shift-selective imaging (product
19F selective) produced the clear signal for probe 1 only upon
incubation with hMAO-A (Figure 4d, left). The presence of an
hMAO-A inhibitor completely suppressed such signal (Figure 4d,
right). These results show clearly that the designed probe 1 func-
tions as the first signal-switching 19F MRI probe for the specific
detection and imaging of hMAO-A activity.
NEXT Program and Grant-in-Aid No. 22685018 from JSPS,
Japan, and partly by the Innovative Techno-Hub for Integrated
Medical Bio-Imaging Project of the Special Coordination Funds
for Promoting Science and Technology from MEXT, Japan.
’ REFERENCES
(1) For reviews on monoamine oxidase, see, for example: (a)
Berry, M. D.; Juorio, A. V.; Paterson, I. A. Prog. Neurobiol. 1994,
42, 375–391. (b) Youdim, M. B.; Edmondson, D.; Tipton, K. F. Nat.
Rev. Neurosci. 2006, 7, 295–309. (c) Edmondson, D. E.; Binda, C.;
Wang, J.; Upadhyay, A. K.; Mattevi, A. Biochemistry 2009, 48,
4220–4230.
(2) (a) Meyer, J. H.; Boovariwala, A.; Sagrati, S.; Hussey, D.; Garcia,
A.; Young, T; Praschak-Rieder, N.; Wilson, A. A.; Houle, S. Arch. Gen.
Psychiatry 2006, 63, 1209–1216. (b) Robinson, D. S. Primary Psychiatry
2007, 14, 32–34.
In conclusion, we report a strategy for designing hMAO-A-
specific 19F MRI probe. The designed MRI probe has at least two
distinct advantages. The first is specificity. On the basis of the
o/p-substituted phenol hypothesis, our MRI probe achieved
excellent discrimination of hMAO-A from its isoform hMAO-B.
hMAO-A and -B play different physiological roles. In particular,
hMAO-A, the focus of the present work, is considered one
possible enzyme involved in the pathogenesis of psychiatric
disorders.2 In this sense, the specificity shown in this work is
essential for revealing the true physiological and pathological
roles of hMAO-A activity. The second advantage is the applic-
ability to 19F MRI. The rationally designed MRI probe achieved
sufficient 19F chemical shift change upon selective reaction with
hMAO-A. These results allow us to conclude that the designed
“smart” molecule functions as the first chemical shift-switching
19F MRI probe for the specific detection and imaging of hMAO-
A activity. In addition to the high performance of the present
probe itself, the design strategy in this report may provide the
basis for creating new hMAO-A- or -B-specific MRI probes.
Judging from the observation that the related 19F-substituted
phenol-type probes could be used successfully for in vivo 19F
NMR analysis of the reporter protein activity (e.g., 0.012À0.026
mmol of probes in the mouse experiments),9 the present probe
seems to have potential for in vivo applications. However,
practical applications must await further experiments to improve
the probe in terms of the enzymatic turnover rates and pharma-
cokinetic profiles. 19F MR spectroscopy in a much larger voxel
might also be a promising approach to overcome the intrinsic low
sensitivity of the MR probe. In the future, the combination with
PET studies should provide a more precise understanding of
MAO activity.12 Further work is in progress along these lines.
(3) (a) Krajl, M. Biochem. Pharmacol. 1965, 14, 1684–1686. (b)
Zhou, J. J. P.; Zhong, B.; Silverman, R. B. Anal. Biochem. 1996,
234, 9–12. (c) Zhou, M.; Panchuk-Voloshina, N. Anal. Biochem.
1997, 253, 169–174. (d) Chen, G.; Yee, D. J.; Gubernator, N. G.;
Sames, D. J. Am. Chem. Soc. 2005, 127, 4544–4545. (e) Albers, A. E.;
Rawls, K. A.; Chang, C. J. Chem. Commun. 2007, 4647–4649. (f) Aw,
J.; Shao, Q.; Yang, Y.; Jiang, T.; Ang, C.; Xing, B. Chem. Asian J. 2010,
5, 1317–1321.
(4) Zhou, W.; Valley, M. P.; Shultz, J.; Hawkins, E. M.; Bernad, L.;
Good, T.; Good, D.; Riss, T. L.; Klaubert, D. H.; Wood, K. V. J. Am.
Chem. Soc. 2006, 128, 3122–3123.
(5) (a) Tabor, C. W.; Tabor, H.; Rosenthal, S. M. J. Biol. Chem. 1954,
208, 645–661. (b) Weissbach, H.; Smith, T. E.; Daly, J. W.; Witkop, B.;
Udenfriend, S. J. Biol. Chem. 1960, 235, 1160–1163. (c) Houslay, M. D.;
Tipton, K. F. Biochem. J. 1974, 139, 645–652. (d) Flaherty, P.;
Castagnoli, K.; Wang, Y.-X.; Castagnoli, N., Jr. J. Med. Chem. 1996,
39, 4756–4761. (e) Bissel, P.; Bigley, M. C.; Castagnoli, K.; Castagnoli,
N., Jr. Bioorg. Med. Chem. 2002, 10, 3031–3041.
(6) (a) Cobb, S. L.; Murphy, C. D. J. Fluorine Chem. 2009,
130, 132–143. (b) Yu, J.-X.; Kodibagkar, V. D.; Cui, W.; Mason, R. P.
Curr. Med. Chem. 2005, 12, 819–848.
(7) (a) Costa, J. L.; Dobson, C. M.; Fay, D. D.; Kirk, K. L.; Poulsen,
F. M.; Valeri, C. R.; Vecchione, J. J. FEBS Lett. 1981, 136, 325–328.
(b) Diffley, D. M.; Costa, J. L.; Sokoloski, E. A.; Chiueh, C. C.; Kirk,
K. L.; Creveling, C. R. Biochem. Biophys. Res. Commun. 1983, 110,
740–745.
(8) For recent examples of 19F NMR probes for enzymatic
analyses, see: (a) Mizukami, S.; Takikawa, R.; Sugihara, F.; Hori,
Y.; Tochio, H.; W€alchli, M.; Shirakawa, M.; Kikuchi, K. J. Am. Chem.
Soc. 2008, 130, 794–795. (b) Stockman, B. J. J. Am. Chem. Soc. 2008,
130, 5870–5871. (c) Tanabe, K.; Harada, H.; Narazaki, M.; Tanaka,
K.; Inafuku, K.; Komatsu, H.; Ito, T.; Yamada, H.; Chujyo, Y.;
Matsuda, T.; Hiraoka, M.; Nishimoto, S. J. Am. Chem. Soc. 2009,
131, 15982–15983. (d) Takaoka, Y.; Sakamoto, T.; Tsukiji, S.;
Narazaki, M.; Matsuda, T.; Tochio, H.; Shirakawa, M.; Hamachi, I.
Nat. Chem. 2009, 1, 557–561.
’ ASSOCIATED CONTENT
(9) (a) Cui, W.; Otten, P.; Li, Y.; Koeneman, K. S.; Yu, J.; Mason,
R. P. Magn. Reson. Med. 2004, 51, 616–620. (b) Yu, J.; Mason, R. P.
J. Med. Chem. 2006, 49, 1991–1999. (c) Liu, L.; Kodibagkar, V. D.; Yu,
J.-X.; Mason, R. P. FASEB J. 2007, 21, 2014–2019. (d) Yu, J.-X.;
Kodibagkar, V. D.; Liu, L.; Mason, R. P. NMR Biomed. 2008,
21, 704–712.
S
Supporting Information. Methods and Figures S1 and
b
S2. This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
(10) Colibus, L. D.; Li, M.; Binda, C.; Lustig, A.; Edmondson, D. E.;
Mattevi, A. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 12684–12689.
(11) Faint artifacts were observed in Figure 4b because of the
excessive brightness of the samples.
Corresponding Author
ssando@ifrc.kyushu-u.ac.jp
(12) Arnett, C. D.; Fowler, J. S.; MacGregor, R. R.; Schlyer, D. J.;
Wolf, A. P.; L^angst€orn, B.; Halldin, C. J. Neurochem. 1987, 2,
522–527.
’ ACKNOWLEDGMENT
We thank Prof. Masahiro Shirakawa of Kyoto University for
support on NMR measurements. This work was supported by
14211
dx.doi.org/10.1021/ja2057506 |J. Am. Chem. Soc. 2011, 133, 14208–14211