Negative-ion electron capture dissociation: Radical-driven fragmentation of charge-increased gaseous peptide anions

Article abstract of DOI:10.1021/ja207736y

The generation of gaseous polyanions with a Coulomb barrier has attracted attention as exemplified by previous studies of fullerene dianions. However, this phenomenon has not been reported for biological anions. By contrast, electron attachment to multiply charged peptide and protein cations has seen a surge of interest due to the high utility for tandem mass spectrometry (MS/MS). Electron capture dissociation (ECD) and electron transfer dissociation (ETD) involve radical-driven fragmentation of charge-reduced peptide/protein cations to yield N-C_α backbone bond cleavage, resulting in predictable c′/z^?-type product ions without loss of labile post-translational modifications (PTMs). However, acidic peptides, e.g., with biologically important PTMs such as phosphorylation and sulfonation, are difficult to multiply charge in positive ion mode and show improved ionization in negative-ion mode. We found that peptide anions ([M - nH]^n-, n ≥ 1) can capture electrons within a rather narrow energy range (～3.5-6.5 eV), resulting in charge-increased radical intermediates that undergo dissociation analogous to that in ECD/ETD. Gas-phase zwitterionic structures appear to play an important role in this novel MS/MS technique, negative-ion electron capture dissociation (niECD).

Full text of DOI:10.1021/ja207736y

COMMUNICATION
pubs.acs.org/JACS
Negative-Ion Electron Capture Dissociation: Radical-Driven
Fragmentation of Charge-Increased Gaseous Peptide Anions
ꢀ
Hyun Ju Yoo,†,§,^ Ning Wang,†,§ Shuyi Zhuang,†,‡ Hangtian Song,† and Kristina Hakansson*,†
†Department of Chemistry, University of Michigan, Ann Arbor, Michigan 48109-1055, United States
‡Department of Chemistry, Tsinghua University, Beijing 100084, China
S Supporting Information
b
charged cations is challenging for acidic analytes, including
ABSTRACT: The generation of gaseous polyanions with a
Coulomb barrier has attracted attention as exemplified by
previous studies of fullerene dianions. However, this phe-
nomenon has not been reported for biological anions. By
contrast, electron attachment to multiply charged peptide
and protein cations has seen a surge of interest due to the
high utility for tandem mass spectrometry (MS/MS).
Electron capture dissociation (ECD) and electron transfer
dissociation (ETD) involve radical-driven fragmentation of
charge-reduced peptide/protein cations to yield NꢁCα
backbone bond cleavage, resulting in predictable c0/z•-type
product ions without loss of labile post-translational mod-
ifications (PTMs). However, acidic peptides, e.g., with
biologically important PTMs such as phosphorylation and
sulfonation, are difficult to multiply charge in positive ion
mode and show improved ionization in negative-ion mode.
We found that peptide anions ([M ꢁ nH]nꢁ, n g 1) can
capture electrons within a rather narrow energy range
(∼3.5ꢁ6.5 eV), resulting in charge-increased radical inter-
mediates that undergo dissociation analogous to that in
ECD/ETD. Gas-phase zwitterionic structures appear to
play an important role in this novel MS/MS technique,
negative-ion electron capture dissociation (niECD).
peptides with important PTMs such as phosphorylation and
sulfonation. Thus, alternative negative-ion MS/MS techniques
are desired. CAD of peptide anions typically results in PTM loss,
similar to cation CAD. Further, backbone fragmentation in
negative-ion CAD is more complex than in positive-ion mode and
not predictable.4 Electron-based techniques operating in negative-
ion mode include electron detachment dissociation (EDD)5aꢁc and
negative electron-transfer dissociation (NETD).6a,6b The former
technique has low fragmentation efficiency, and the latter can result
in PTM loss due to energy release from charge reduction. Both
EDD and NETD yield backbone a•- and x-type product ions but
involve structurally uninformative neutral losses as major fragmen-
tation pathways. In addition, both techniques require multiply
charged anions as precursors. Metastable atom-activated dissocia-
tion,7a,b also believed to involve radical-driven dissociation, was
recently shown to yield fragmentation complementary to CAD,
ECD, and EDD, with little PTM loss for peptide anions.8
Electron capture by anionic gaseous peptides appears unlikely
due to Coulomb repulsion. However, previous work has shown
attachment of 2ꢁ3 eV electrons to singly charged fullerene
anions to form dianions in a Fourier transform ion cyclotron
resonance (FT-ICR) mass spectrometer.9 Electron transfer to
unmodified10 and fluorinated11 fullerene anions was also ob-
served in high-energy (keV) collisions with atomic and molecular
targets. We argued that such a phenomenon may be feasible for
peptide anions at a certain mass-to-charge (m/z) ratio and an
appropriate electron energy. To test this hypothesis, we started
with coumarin-tagged peptides, based on work by O’Connor
et al. who showed that coumarin acts as a radical trap in
conventional cation ECD.12 After careful optimization of the
electron energy, we observed abundant charge-increased radical
species, [M + coumarin ꢁ H]2ꢁ•, generated from capture of
∼4.5 eV electrons (corresponding to a cathode bias voltage of
6 V, see Figure S1) by singly deprotonated coumarin-tagged
peptides following 20 s electron irradiation (Figures 1A and S2).
The charge-increased radical species from angiotensin I
(Figure 1A) was isolated in the ICR cell to verify that this product
is not an artifact at twice the precursor ion ICR frequency
(Figure S3). These data demonstrate the feasibility of electron
capture by peptide anions, but the generated doubly charged radical
anions appeared stable to further dissociation, consistent with
the previously observed behavior of coumarin-tagged peptides12
and peptides containing other electron predators13a,b in
as-phase ionꢁelectron and ionꢁion reactions are gaining
G
popularity for peptide activation in tandem mass spectro-
metry (MS/MS). Electron capture dissociation (ECD)1 and
electron transfer dissociation (ETD)2 are powerful alternatives
to collision-activated dissociation (CAD). Fragmentation pat-
terns observed in electron-mediated MS/MS are complementary
to those observed in CAD, frequently providing more extensive
peptide sequence information and, importantly, not involving loss
of labile post-translational modifications (PTMs). Thus PTM
sites can be determined, which is often challenging with CAD.
More recently, electron ionization and subsequent extensive dis-
sociation (electron ionization dissociation, EID) has been reported
following irradiation of [M + nH]n+ (n g 1) peptide cations with
fast electrons (>20 eV).3 Such irradiation causes double ionization
to [M+nH](n+2)+ followed by electron capture to form electro-
nically excited [M + nH](n+1)+•* ions, which dissociate via both side-
chain losses and backbone fragmentation.
ECD, ETD, and EID all involve positively charged precursor
ions with at least two charges for ECD/ETD because capture/
transfer of an electron reduces total charge by 1, and mass
spectrometers cannot detect neutrals. Generation of multiply
Received: August 16, 2011
Published: September 26, 2011
r
2011 American Chemical Society
16790
dx.doi.org/10.1021/ja207736y J. Am. Chem. Soc. 2011, 133, 16790–16793
|

Journal of the American Chemical Society
COMMUNICATION
Figure 2. niECD of α-casein tryptic phosphopeptides. (A) 10 s irradiation
of a singly deprotonated, singly phosphorylated peptide (∼4.5 eV electrons,
10 scans). (B) 20 s irradiation of a doubly deprotonated, doubly phos-
phorylated phosphopeptide (∼5.5 eV electrons, 10 scans).
electron energy range (∼3.5ꢁ6.5 eV) appears acceptable for
niECD (Figure 1C).
Figure 1. (A) Electron irradiation (∼4.5 eV electrons, 20 s, 1 scan) of
singly deprotonated coumarin-tagged angiotensin I. (B) niECD of singly
deprotonated un-derivatized angiotensin I (∼4.5 eV electrons, 20 s
irradiation, 5 scans). (C) Abundance change of the charge-increased
[M ꢁ H]2ꢁ• species as a function of electron energy. (D) IRMPD MS3
(10.6 μm, 300 ms, 7.5 W, 20 scans) of the in-cell-isolated [M ꢁ H]2ꢁ•
species generated upon electron capture by un-derivatized angiotensin I.
Charge-increased product/precursor ions are marked in red. ν3 = third
harmonic; * = electronic noise.
To further investigate the observed gas-phase chemi-
stry, the charge-increased species, [M ꢁ H]2ꢁ•, generated from
electron irradiation of singly deprotonated un-derivatized angio-
tensin I was isolated in the ICR cell and activated in an MS3
experiment via IRMPD (Figure 1D). Two major backbone
fragments are observed in the form of complementary c3 and
z7 ions. The different outcomes in direct niECD (Figure 1B) and
MS3 involving electron irradiation followed by IRMPD of the
isolated, charge-increased [M ꢁ H]2ꢁ• species are likely due to
the different time scales of the two experiments. Lin et al.
previously demonstrated that several radical intermediates with
different lifetimes exist in conventional cation ECD.17 Never-
theless, c/z-type ions were observed in both MS2 and MS3. A
comparison between MS3 of the charge-increased [M ꢁ H]2ꢁ•
radical species and direct IRMPD (MS2) of even-electron
[M ꢁ 2H]2ꢁ precursor ions is shown in Figure S5 for the peptide
H-KRSpYEEHIP-OH. Very different product ion spectra result,
with solely z-type ions observed for radical precursors and mainly
b- and y-type ions for even-electron precursor ions.
niECD of α-casein tryptic phosphopeptides is shown in
Figure 2. For a singly deprotonated serine-phosphorylated pep-
tide (Figure 2A), ∼4.5 eV electron irradiation yields an ammonia-
deficient charge-increased radical, [M ꢁ NH3 ꢁ H]2ꢁ•, as the
major product along with three doubly charged and many sin-
gly charged c and z ions. For doubly charged precursor ions
(e.g., the doubly deprotonated and doubly phosphorylated α-
casein tryptic peptide shown in Figure 2B), the optimum niECD
electron energy is slightly higher than for singly charged pre-
cursor ions, ∼5.5 rather than ∼4.5 eV, consistent with increased
Coulomb repulsion. The fragmentation efficiency is lower for
doubly charged precursor ions; however, a charge-increased
conventional ECD, and with the previously observed fullerene
dianions.9 Further activation of the generated [M + coumarin ꢁ
H]2ꢁ• radical species through infrared multiphoton dissociation
(IRMPD14) mainly resulted in ejection of small, structurally
uninformative neutrals (Figure S3, inset).
Following our discovery that ∼4.5 eV electrons can be
captured by coumarin-tagged peptide anions, we applied electron
irradiation to unmodified peptide anions. Figure 1B shows
electron irradiation of singly deprotonated angiotensin I with-
out a coumarin tag. Similar to the coumarin-tagged species
(Figure 1A), a charge-increased radical anion, [M ꢁ H]2ꢁ•, is
observed, but, in contrast to the coumarin-tagged species, several
c0- and z•-type fragments (Zubarev nomenclature)15 from back-
bone NꢁCα bond cleavage are also detected. We termed
this phenomenon negative-ion electron capture dissociation
(niECD). Remarkably, a doubly charged c09 ion is observed from
the singly charged precursor ion. Electron-induced dissociation
at higher electron energy16 (∼9.5 eV, Figure S4A) does not yield
any charge-increased product ions or c/z-type fragments, and
such fragments are also absent in CAD of the same species
(Figure S4B), indicating that niECD proceeds through a unique
mechanism related to that of conventional peptide ECD/ETD
(which also yield c/z-type product ions). A rather narrow
16791
dx.doi.org/10.1021/ja207736y |J. Am. Chem. Soc. 2011, 133, 16790–16793

Journal of the American Chemical Society
COMMUNICATION
Table 1. Comparison of niECD and CAD for Phosphopep-
tide Anionsa
a Backbone NꢁCα bond cleavages to yield c0/z• ions are indicated with
red lines, and backbone amide bond cleavages to yield b/y0 ions are
indicated with green lines. Dashed lines indicate accompanying phos-
phate loss. Lack of indicated fragments in CAD is due to extensive
neutral losses (e.g., HPO3, H3PO4, and H2O).
anion) had the sequence H-RRApSVA-OH. This resistance to
niECD is likely due to the smaller molecular weight and thus
decreased favorability for accommodating two negative charges
in the gas phase.
The niECD outcome of several unmodified peptides is
summarized in Table S1. In contrast to phosphopeptides
(Figure 2, Table 1), for which all but one short peptide showed
extensive fragmentation in niECD, several singly deprotonated
non-phosphopeptides did not capture electrons, including the
larger (>1 kDa) peptides cholecystokinin, neurokinin B, sub-
stance P-OH, neuromedin C, and neuromedin B. One common
characteristic of these five peptides is a lack of either strongly
basic or strongly acidic residues, thus reducing the probability of
gas-phase zwitterionic structures. In addition, previous work has
shown that such structures are favored for phosphopeptides,18a,b
which also undergo favorable niECD (Figure 2, Table 1). These
observations, along with the striking similarity of niECD spectra
to cation ECD/ETD spectra, suggest that zwitterionic structures
may play an important role for successful niECD, with electron
capture either occurring at or being directed by the positively
charged site.19aꢁc Work by Vasil’ev et al. involving electron
capture by neutral gaseous peptides20 showed somewhat different
product ion spectra, with a larger variety of fragment types
compared to niECD, further suggesting that charged sites may
play a role in niECD. Furthermore, recent computational work
proposes that singly deprotonated angiotensin II is zwitterionic.21
To test this zwitterion hypothesis, we performed several experi-
ments with the goal to either prevent or promote gas-phase peptide
zwitterion formation. Figure 3B shows niECD of N-terminally
acetylated CCKS. This sulfopeptide showed highly favorable niECD
in its unmodified form (Figure 3A); however, it does not contain any
basic residues. Thus, the most basic site is the N-terminus, and
zwitterion formation should be less favorable upon acetylation.
Consistently, niECD efficiency of N-terminally acetylated CCKS is
significantly lower than that of unmodified CCKS (Figure 3A,B).
However, electron capture and fragmentation still occur for the
acetylated species, possibly due to tryptophan protonation.
Figure 3. (A) niECD of sulfonated cholecystokinin (CCKS; ∼4.5 eV
electrons, 20 s, 10 scans). (B) niECD of N-terminally acetylated CCKS
under conditions identical to those in (A). (C) niECD of trimethylam-
monium-derivatized CCK (∼4.5 eV electrons, 10 s, 32 scans). Charge-
increased product ions are highlighted in red. Fragmentation efficiency
was calculated as previously described.3
triply charged radical, [M ꢁ 2H]3ꢁ•, is observed along with four
other charge-increased products and many doubly charged c/z
ions. Phosphate loss is absent for both phosphopeptides.
Figure 3A shows niECD of a tyrosine-sulfonated peptide
(cholecystokinin, CCKS). Sulfonation is even more labile in
the gas phase than phosphorylation and frequently lost in
positive-ion mode, even without ion activation. Thus, negative
mode, in which sulfotyrosine is stable and shows higher ionization
efficiency, is preferred compared to positive mode analysis. EDD
has shown some success for sulfonate localization in sulfopep-
tides;5 however, backbone fragmentation competes with neutral
loss of CO2 and SO3. In niECD (Figure 3A), no sulfonate loss
occurs, and virtually complete sequence coverage is observed.
Additional examples of phosphopeptide niECD and compari-
son to anion CAD are shown in Table 1. In all cases, niECD
provides significantly more extensive peptide sequence coverage
than CAD, and both serine and tyrosine phosphorylation are
retained. The only phosphopeptide we analyzed that did not
undergo niECD (or electron capture by the singly deprotonated
Intriguingly, non-sulfonated CCKS (CCK) does not undergo
niECD (Table S1). Similarly, gas-phase desulfonation of CCKS
via nozzleꢁskimmer dissociation inside the electrospray ion
source eliminates electron capture by the resulting CCK-like
product (Figure S6). Addition of metal ions (Na+, Ca2+, and
Cs+), which may promote zwitterion formation,22 did not enable
electron capture by CCK (Figure S7). Neither did N-terminal
16792
dx.doi.org/10.1021/ja207736y |J. Am. Chem. Soc. 2011, 133, 16790–16793

Journal of the American Chemical Society
COMMUNICATION
tris(2,4,6-trimethoxyphenyl)phosphonium-acetyl (TMPP-Ac)
derivatization, which introduces a fixed positive charge and
thereby forces observed singly charged anions to have two
deprotonation sites (Figure S8). By contrast, introduction of a
fixed positive charge in the form of an N-terminal quaternary
amine did enable niECD of CCK (Figure 3C). Quaternary amine
derivatization also rescued niECD ability of substance P-OH
(Figure S9). However, the presence of the fixed charge site
altered the fragmentation behavior in both cases, similar to
reported behavior of fixed charge-containing peptides in con-
ventional ECD/ETD.23aꢁd The lack of success for TMPP-Ac
derivatization or metal adduction may be explained by effective
shielding of the positively charged site by the aromatic groups
surrounding the phosphonium, or by the peptide carbonyls
wrapping around the metal ion. Thus, the presence of a gas-
phase zwitterion does not appear to be the only criterion for
successful niECD. The particular gas-phase zwitterion structure
is likely also crucial: the influence of peptide gas-phase structure has
been extensively studied in conventional cation ECD and is known
to have a profound influence on fragmentation behavior.24aꢁd
In summary, we show that peptide anions can capture
∼3.5ꢁ6.5 eV electrons, resulting in radical species with in-
creased charge and yielding peptide backbone bond fragmenta-
tion (niECD) analogous to that observed in cation ECD/ETD,
including PTM retention and higher sequence coverage compared
to CAD. Increased charge improves signal-to-noise ratios in FT-ICR
MS because the generated image current is proportional to the
charge state.3 The presence of a coumarin radical trap improved
electron capture efficiency but limited fragmentation, presumably
due to decreased radical mobility. niECD allows localization of
PTMs and de novo sequencing for acidic peptides that show
improved ionization in negative-ion mode compared to positive-
ion mode, e.g., phospho- and sulfopeptides. Further, niECD is
compatible with (but not limited to) singly charged peptides, which
allows coupling with matrix-assisted laser desorption/ionization.
’ REFERENCES
(1) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem.
Soc. 1998, 120, 3265.
(2) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt,
D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528.
(3) Fung, Y. M. E.; Adams, C. M.; Zubarev, R. A. J. Am. Chem. Soc.
2009, 131, 9977.
(4) Bowie, J. H.; Brinkworth, C. S.; Dua, S. Mass Spectrom. Rev. 2002,
21, 87.
(5) (a) Budnik, B. A.; Haselmann, K. F.; Zubarev, R. A. Chem. Phys.
Lett. 2001, 342, 299. (b) Kweon, H. K.; Hakansson, K. J. Proteome Res.
2008, 7, 745. (c) Wolff, J. J.; Chi, L.; Linhardt, R. J.; Amster, I. J. Anal.
Chem. 2007, 79, 2015.
(6) (a) Coon, J. J.; Shabanowitz, J.; Hunt, D. F.; Syka, J. E. P. J. Am. Soc.
Mass Spectrom. 2005, 16, 880. (b) Huzarska, M.; Ugalde, I.; Kaplan, D. A.;
Hartmer, R.; Easterling, M. L.; Polfer, N. C. Anal. Chem. 2010, 82, 2873.
(7) (a) Misharin, A. S.; Silivra, O. A.; Kjeldsen, F.; Zubarev, R. A.
Rapid Commun. Mass Spectrom. 2005, 19, 2163. (b) Berkout, V. D. Anal.
Chem. 2009, 81, 725.
(8) Cook, S. L.; Jackson, G. P. J. Am. Soc. Mass Spectrom. 2011,
22, 1088.
(9) Hartig, J.; Blom, M. N.; Hampe, O.; Kappes, M. M. Int. J. Mass
Spectrom. 2003, 229, 93.
(10) Liu, B.; Hvelplund, P.; Nielsen, S. B.; Tomita, S. Phys. Rev. Lett.
2004, 92, 168301.
(11) Boltalina, O. V.; Hvelplund, P.; Larsen, M. C.; Larsson, M. O.
Phys. Rev. Lett. 1998, 80, 5101.
(12) Belyayev, M. A.; Cournoyer, J. J.; Lin, C.; O’Connor, P. B. J. Am.
Soc. Mass Spectrom. 2006, 17, 1428.
(13) (a) Jones, A. W.; Mikhailov, V. A.; Iniesta, J.; Cooper, H. J.
J. Am. Soc. Mass Spectrom. 2010, 21, 268. (b) Sohn, C. H.; Chung, C. K.;
Yin, S.; Ramachandran, P.; Loo, J. A.; Beauchamp, J. L. J. Am. Chem. Soc.
2009, 131, 5444.
(14) Bomse, D. S.; Woodin, R. L.; Beauchamp, J. L. J. Am. Chem. Soc.
1979, 101, 5503.
(15) Kjeldsen, F.; Haselmann, K.; Budnik, B. A.; Jensen, F.; Zubarev,
R. A. Chem. Phys. Lett. 2002, 356, 201.
(16) Kalli, A.; Grigorean, G.; Hakansson, K. J. Am. Soc. Mass
Spectrom. 2011, DOI: 10.1007/s13361-011-0233-6.
(17) Lin, C.; O’Connor, P. B.; Cournoyer, J. J. J. Am. Soc. Mass
Spectrom. 2006, 17, 1605.
’ ASSOCIATED CONTENT
(18) (a) Creese, A. J.; Cooper, H. J. J. Am. Soc. Mass Spectrom. 2008,
19, 1263. (b) Woods, A. S. J. Proteome Res. 2004, 3, 478.
(19) (a) Simons, J. Chem. Phys. Lett. 2010, 484, 81. (b) Syrstad, E. A.;
Turecek, F. J. Am. Soc. Mass Spectrom. 2005, 16, 208. (c) Zubarev, R. A.;
Kruger, N. A.; Fridriksson, E. K.; Lewis, M. A.; Horn, D. M.; Carpenter,
B. K.; McLafferty, F. W. J. Am. Chem. Soc. 1999, 121, 2857.
(20) Vasil’ev, Y. V.; Figard, B. J.; Morre, J.; Deinzer, M. L. J. Chem.
Phys. 2009, 131, 044317.
(21) Marchese, R.; Grandori, R.; Carloni, P.; Raugei, S. PLoS
Comput. Biol. 2010, 6, e1000775.
(22) Jockusch, R. A.; Price, W. D.; Williams, E. R. J. Phys. Chem. A
1999, 103, 9266.
(23) (a) Xia, Y.; Gunawardena, H. P.; Erickson, D. E.; McLuckey,
S. A. J. Am. Chem. Soc. 2007, 129, 12232. (b) Chamot-Rooke, J.; van der Rest,
G.; Dalleu, A.; Bay, S.; Lemoine, J. J. Am. Soc. Mass Spectrom. 2007, 18, 1405.
(c) Chung, T. W.; Moss, C. L.; Zimnicka, M.; Johnson, R. S.; Moritz, R. L.;
Turecek, F. J. Am. Soc. Mass Spectrom. 2011, 22, 13. (d) Li, X. J.; Cournoyer,
J. J.; Lin, C.; O’Connor, P. B. J. Am. Soc. Mass Spectrom. 2008, 19, 1514.
(24) (a) Mihalca, R.; Kleinnijenhuis, A. J.; McDonnell, L. A.; Heck,
A. J. R.; Heeren, R. M. A. J. Am. Soc. Mass Spectrom. 2004, 15, 1869. (b)
Robinson, E. W.; Leib, R. D.; Williams, E. R. J. Am. Soc. Mass Spectrom.
2006, 17, 1469. (c) Ben Hamidane, H.; He, H.; Tsybin, O. Y.; Emmett,
M. R.; Hendrickson, C. L.; Marshall, A. G.; Tsybin, Y. O. J. Am. Soc. Mass
Spectrom. 2009, 20, 1182. (d) Breuker, K.; Oh, H. B.; Lin, C.; Carpenter,
B. K.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 14011.
S
Supporting Information. Methods figures, and table.
b
This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
kicki@umich.edu
Present Addresses
^Analytical Chemistry Core Lab, ASAN Medical Center, 88,
Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, Korea
Author Contributions
§These authors contributed equally to this work.
’ ACKNOWLEDGMENT
This work was supported by an NSF Career Award (CHE-05-
47699). H.J.Y. and H.S. were partially supported by George
Ashworth Analytical Chemistry Fellowships. S.Z. was supported
by a Summer Undergraduate Research Exchange Program spon-
sored by Pfizer Global R&D and Tsinghua University School of
Life Sciences. R. A. Zubarev and K. E. Hersberger are thanked for
valuable discussions.
16793
dx.doi.org/10.1021/ja207736y |J. Am. Chem. Soc. 2011, 133, 16790–16793