Oxidative modification of native protein residues using cerium(IV) ammonium nitrate

Article abstract of DOI:10.1021/ja206324q

A new protein modification strategy has been developed that is based on an oxidative coupling reaction that targets electron-rich amino acids. This strategy relies on cerium(IV) ammonium nitrate (CAN) as an oxidation reagent and results in the coupling of tyrosine and tryptophan residues to phenylene diamine and anisidine derivatives. The methodology was first identified and characterized on peptides and small molecules, and was subsequently adapted for protein modification by determining appropriate buffer conditions. Using the optimized procedure, native and introduced solvent-accessible residues on proteins were selectively modified with polyethylene glycol (PEG) and small peptides. This unprecedented bioconjugation strategy targets these under-utilized amino acids with excellent chemoselectivity and affords good-to-high yields using low concentrations of the oxidant and coupling partners, short reaction times, and mild conditions.

Full text of DOI:10.1021/ja206324q

ARTICLE
pubs.acs.org/JACS
Oxidative Modification of Native Protein Residues Using Cerium(IV)
Ammonium Nitrate
Kristen L. Seim,† Allie C. Obermeyer,† and Matthew B. Francis*,†,‡
†Department of Chemistry, University of California, Berkeley, California 94720-1460, United States
‡Materials Sciences Division, Lawrence Berkeley National Laboratories, Berkeley, California 94720-1460, United States
S Supporting Information
b
ABSTRACT: A new protein modification strategy has been
developed that is based on an oxidative coupling reaction that
targets electron-rich amino acids. This strategy relies on cerium-
(IV) ammonium nitrate (CAN) as an oxidation reagent and
results in the coupling of tyrosine and tryptophan residues to
phenylene diamine and anisidine derivatives. The methodology
was first identified and characterized on peptides and small
molecules, and was subsequently adapted for protein modification by determining appropriate buffer conditions. Using the
optimized procedure, native and introduced solvent-accessible residues on proteins were selectively modified with polyethylene
glycol (PEG) and small peptides. This unprecedented bioconjugation strategy targets these under-utilized amino acids with
excellent chemoselectivity and affords good-to-high yields using low concentrations of the oxidant and coupling partners, short
reaction times, and mild conditions.
’ INTRODUCTION
amino acids that can be introduced without specialized expres-
sion systems.
The development of reliable, chemoselective reactions for
protein modification is an important pursuit in the field of
chemical biology, and these methods provide much-needed tools
for the creation of protein-based materials.1 Through the attach-
ment of synthetic molecules to specific locations on protein surfaces,
biomolecular structure and function can be both elucidated and
exploited in a diverse number of contexts. To be useful in most
cases, bioconjugation reactions must create precise, well-defined
linkages without the use of protecting groups. In addition to the
high degree of chemoselectivity, these reactions must proceed in
aqueous solution under mild pH and temperature conditions and
at low substrate concentrations. Although traditional bioconju-
gation strategies that target abundant amino acid side chains
(such as lysine and aspartic and glutamic acids) are useful for
nonspecific labeling, it is often difficult to control the number and
locations of the modifications introduced at these sites.1 When
site specificity is required, the most utilized and reliable biocon-
jugation strategies target cysteine residues. These are particularly
effective due to the low abundance2 and nucleophilic nature of
reduced thiolates on protein surfaces. However, there are many
cases in which the introduction of a unique cysteine is incon-
venient or difficult without disrupting protein function. Addi-
tionally, a growing number of protein-based materials require
modification in two distinct locations, establishing a need for
methods that can be combined with cysteine-targeting strategies
to introduce multiple synthetic groups.3 As one option, recent
work has addressed these needs by introducing artificial amino
acids and modifying the introduced functionalities with new
protein-compatible chemistry.4 While these approaches are
very promising, there are still advantages to targeting native
As an alternative solution to this problem, our group has
explored the development of site-selective bioconjugation reac-
tions that target the aromatic electron-rich amino acids tyrosine
and tryptophan.5À8 These amino acids occur with intermediate
to low frequency on native protein surfaces2 and are often partially
or completely buried due to the amphiphilic nature of the phenolic
and indole groups. Considering the complexities of these envir-
onments, it is therefore likely that individual, solvent-accessible
tyrosine and tryptophan residues could be targeted though careful
reagent design and selective optimization of reaction conditions.
These amino acids, like cysteine, could also provide sites for
selective modification when introduced on protein surfaces using
site-directed mutagenesis. Despite their potential utility, broadly
applicable, highly specific chemical reactions that target tyrosine
and tryptophan have traditionally been lacking. Recent work has
expanded the number of techniques for the modification of these
amino acids, including the utilization of rhodium carbenoids5,9,10
and malondialdehyde derivatives11À13 for the alkylation of try-
ptophan and the use of Mannich-type additions,7 π-allylpalla-
dium complexes,8 acyclic diazodicarboxamide reagents,14 and
diazonium salts6 for the alkylation of tyrosine. Although the
development of these techniques represents significant improve-
ments in tryptophan- and tyrosine-targeting bioconjugation
strategies, many of these methods would benefit from improved
yields and lower levels of cross-reactivity.
Received: July 14, 2011
Published: October 03, 2011
r
2011 American Chemical Society
16970
dx.doi.org/10.1021/ja206324q J. Am. Chem. Soc. 2011, 133, 16970–16976
|

Journal of the American Chemical Society
ARTICLE
In order to develop new bioconjugation strategies that target
these amino acids, our current studies have focused on taking
advantage of the chemical properties of the electron-rich aro-
matic rings and their susceptibility to oxidation. Tyrosine and
tryptophan residues have been shown to be preferred targets of
reactive oxygen species, and several studies have documented the
structures and consequences of oxidized residues in proteins.15À17
Additionally, oxidized radical species of amino acids in biology
are not only common, but are also crucial for many significant
energy conversion processes.18À21 The one-electron oxidation
chemistry of tyrosine has been previously exploited by Kodadek
and co-workers to cross-link proteins using metal-catalyzed
oxidations.22À24 Although the majority of these coupling reac-
tions are hypothesized to occur through the addition of tyrosyl
radicals to adjacent tyrosine residues, oxidized tryptophan radi-
cals are also thought to play a role in some cases. One might
expect that similarly generated amino acid radicals could add to
electron-rich synthetic components, or conversely, the closed-
shell forms of the amino acids could couple to free radicals in
solution. Although the oxidation of other amino acids is well-
documented and could pose a potential selectivity problem, we
anticipated that reagents and conditions could be optimized for
controlled and selective modifications that are limited to tyrosine
and tryptophan residues.
In this article we describe an unprecedented oxidative cou-
pling strategy that can selectively modify tyrosine and tryptophan
residues using cerium(IV) ammonium nitrate (CAN) as a one-
electron oxidant. This oxidative method was found to couple
electron-rich aniline derivatives directly to tyrosine and trypto-
phan residues. This method was adapted for the chemoselective
modification of proteins and was optimized to proceed with high
yields at neutral pH and low substrate concentrations. The strategy
has been used to modify both native and introduced residues on
proteins with polyethylene glycol (PEG) and small peptides. The
reaction was also used in conjunction with cysteine alkylation
to doubly modify viral capsid proteins with both targeting and
imaging functionalities.
Figure 1. Modification of aromatic residues (W and Y) on (a) peptides
using (b) phenylene diamine (1) and anisidine (2) derivatives in the
presence of cerium(IV) ammonium nitrate (CAN). Reaction condi-
tions: 1.5 mM CAN, 500 μM 1 or 2, 100 μM peptide, 10 mM HEPES
buffer, pH 4.5, 1 h. The reaction mixtures in (c) and (d) were analyzed
using MALDI-TOF MS.
identified involved N-acyl phenylene diamine derivative 1, which
was found to modify melittin and angiotensin a single time in the
presence of CAN, Figure 1c.25 MS/MS analysis of the modified
peptides confirmed the chemoselectivity of these conditions for
the modification of the tryptophan and tyrosine residues of
melittin and angiotensin, respectively (Figures S1 and S2 in the
Supporting Information [SI]). Alkylated anisidine 2 was also
found to modify angiotensin in the presence of CAN, but this
combination did not lead to product formation for melittin,
Figure 1d.26 This result was interpreted to indicate that the anisidine
addition was selective for tyrosine residues. MS/MS analysis
of the modified angiotensin was used to confirm this chemo-
selectivity (Figure S3 in the SI). In all cases, the mass changes
observed suggested an addition of the small molecules to the
indole and phenol rings with the loss of two hydrogen atoms.
As a result of the promising levels of modification and clean
reactivity observed with these reagents, the phenylene diamine
and anisidine derivatives were carried on for further study.
To be broadly applicable for protein modification, bioconju-
gation reactions must proceed using buffered conditions that are
close to neutral pH. However, the simple addition of CAN to
unbuffered aqueous solution leads to significant acidification,27
with a 1.5 mM solution of CAN in unbuffered H2O reaching pH
∼2À3. To counteract this unfavorable pH change, a series of
buffers was screened, with the goal of maintaining neutral pH
without precipitating or deactivating the Ce(IV) reagent.28 These
screens were performed in triplicate on melittin and angiotensin,
and the relative product yields were determined using MALDI-
TOF MS (see Figure S5 in the SI for response curves for
quantitation by MALDI-TOF MS). The results are tabulated
’ RESULTS AND DISCUSSION
To determine compatible combinations of reactants and oxi-
dants, a variety of candidates was screened for the modification of
peptides in aqueous solution. To identify any reaction products
that had formed, the reaction mixtures were analyzed using
MALDI-TOF MS. The peptides used for screening were melittin
as a tryptophan-containing peptide and angiotensin as a tyrosine-
containing peptide, Figure 1a. These were combined with a series
of electron-rich aromatic compounds and oxidants, such as
ammonium persulfate, N-bromosuccinimide, N-chlorosuccinimide,
Oxone, sodium periodate, potassium ferricyanide, and cerium-
(IV) ammonium nitrate (CAN). Initial screens identified CAN
as a promising oxidant for the modification of these peptides, as
this oxidant was found to efficiently couple aniline derivatives to
one or both peptides. Anilines bearing additional electron-donating
substituents (such as methoxy- and acetamido- groups) led to the
highest levels of reactivity, while nitro- and halogen-substituted
anilines were unreactive. Primary anilines generally lead to over-
modification and complex product mixtures, likely due to the
self-condensation reactions that occur with these species under
oxidizing conditions. In contrast, N,N-dialkyl anilines bearing
additional electron-donating substituents resulted in single addi-
tions to the peptides. One of the most promising combinations
16971
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976

Journal of the American Chemical Society
ARTICLE
Figure 2. Effects of buffer composition on pH stability and product
formation. Reaction conditions: 1.5 mM CAN, 500 μM 1 or 2, 100 μM
peptide, RT, 1 h. Each reaction was repeated in triplicate and product
ratios were determined using MALDI-TOF MS. Average product ratios
are reported.
in Figure 2. Although a number of buffers were found to be
compatible with the reaction, reactivity was typically lost as the
strength of the buffers was increased to maintain mild pH condi-
tions. For example, angiotensin could be modified in near-quan-
titative yield in 10 mM TRIS buffer (entry 10). However, the pH
dropped to ∼4 when the CAN was added at 1.5 mM (for
representative MALDI-TOF MS spectra, see Figure S6 in the SI).
Although most peptides can withstand these acidic conditions,
proteins are more likely to denature at pH 4. For protein sub-
strates, 50 mM bis(TRIS) was likely to be the optimal buffer, as it
was able to maintain a pH of 6.0 with a minimal loss in reactivity
(entry 14). In this buffer, a 100 μM solution of melittin was
modified with a 50% conversion in 1 h using 1.5 mM CAN and
500 μM 1. Similarly, a 100 μM solution of angiotensin was modi-
fied with 60À80% conversion using 1.5 mM CAN and 500 μM
1 or 2 and the same reaction time (for representative MALDI-
TOF MS spectra, see Figure S7 in the SI).
In order to characterize the reactivity observed on peptides,
the structures of the modified amino acids were determined. To
identify these modification products, we designed water-soluble,
small-molecule tryptophan and tyrosine analogs that could be
exposed to the reactive conditions used on peptides. The syn-
thesized amino acid analogs (3 and 5) were subjected to the
oxidative conditions in 50 mM bis(TRIS) buffer, pH 6.0. Any
possible products were purified by flash chromatography and
were analyzed by one-dimensional and two-dimensional NMR,
as well as high-resolution mass spectrometry.
Figure 3. Reaction products obtained for small molecule (a) trypto-
phan and (b, c) tyrosine analogs. Product structures were determined
using NOESY and HMBC NMR analyses (for spectra see SI).
bond on C-2 of the indole ring (Figure 3a; for 2D NMR spectra,
see Figures S8 and S9 in the SI). High-resolution MS confirmed
that the mass of the isolated small molecule corresponded to the
structure that was identified using NMR, as well as to the mass
change seen for melittin under analogous reaction conditions.
The reaction of 5 with 1 in the presence of CAN resulted in a
60:40 ratio of two isomeric products that could be separated by
flash chromatography. High-resolution MS confirmed that the
mass of both products corresponded to the mass changes observed
for angiotensin under similar reactive conditions. Two-dimen-
sional NMR analyses were used to identify the structural differences
between these two products. The major product (6) was found
to contain a bond between the tyrosine oxygen and an aromatic
carbon on the phenylene diamine coupling partner. This bond
was found to be ortho to the alkylated nitrogen on the phenylene
diamine ring despite the steric demands of this substitution
pattern. The minor product (7) was determined to be the result
of carbonÀcarbon bond formation ortho to the phenolic oxygen
of the tyrosine. This bond was also found to be located ortho to
the alkylated nitrogen on the aromatic ring of the phenylene
diamine coupling partner (Figure 3b; for 2D NMR spectra, see
Figures S10ÀS13 in the SI).
The coupling of 3 to 1 in the presence of CAN resulted in a
single product (4) that could be isolated by flash chromatogra-
phy. Two dimensional (2D) NMR analyses indicated that this
compound resulted from an alkylation of the amide nitrogen of
the phenylene diamine coupling partner, forming a carbonÀnitrogen
When reacted with 2 in the presence of CAN, tyrosine
analog 5 formed a 85:15 ratio of two isomeric products that
could be separated by flash chromotography. These products
were found to be structurally analogous to those identified in the
16972
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976

Journal of the American Chemical Society
ARTICLE
phenylene diamine reaction. The major product resulted from
the formation of a similar carbonÀoxygen bond, while the minor
product resulted from the formation of an aromatic carbonÀ
carbon bond (Figure 3c; for 2D NMR spectra see Figures S14À17
in the SI). As was the case for phenylene diamine 1, each bond
was formed at the position ortho to the alkylated nitrogen of the
anisidine derivative. Additionally, the mass of the products cor-
responded to the identified structures and to the mass changes
observed for angiotensin. Although the mechanisms for these
modifications are currently under investigation, the formation of
the observed products can be rationalized through radical coupling
pathways (see Figure S18 in the SI for our current mechanistic
hypotheses).
In order to evaluate the identified reactions on protein sub-
strates, new tyrosine and tryptophan residues were introduced to
the external surface of genome-free MS2 capsids. Composed of
180 monomers, MS2 is a porous, hollow structure with icosahe-
dral symmetry.29 After expression in Escherichia coli, the coat
protein monomers spontaneously self-assemble into noninfec-
tious capsids with 32 pores that allow access to the interior cavity.
Although the sequence of wild-type MS2 includes a number of
native tyrosine and tryptophan residues, the majority of these
residues are not solvent-accessible and unlikely to modify with
our coupling conditions. An exception is tyrosine 85, a residue on
the interior surface of MS2 that has been successfully modified
using alternative tyrosine modification strategies.7 Although the
presence of this internal tyrosine could result in double mod-
ification, we expected that an introduced tyrosine on the exterior
surface could be selectively modified using a large molecular
weight coupling partner (such as a modified PEG chain) that
could not diffuse through the ∼2 nm pores. Additionally, if there
were any difference in the solvent accessibilities of the two
tyrosine residues, we might expect that the reaction conditions
could be altered to achieve the selective modification of only the
most solvent-accessible site.
Using site-selective mutagenesis, tyrosine and tryptophan
residues were introduced into the external positions 19 and 15
of the MS2 monomer (Figure 4a). A non-natural p-aminophenol
introduced at position 19 has been previously modified with high
yields and selectivity with an alternative oxidative modification
strategy.4f We therefore expected an introduced tyrosine or
tryptophan at that position to be solvent-accessible and to be
modified with similar success using the new Ce(IV)-based reaction.
In addition, we chose to add the new amino acids at position 15,
which has previously been shown to tolerate the introduction of a
cysteine residue for use in subsequent alkylation reactions.30
Located on the turn of a loop that extended from the external
surface (Figure 4a), a tryptophan or tyrosine residue located at
position 15 was likely to be solvent-accessible and might be
expected to modify more readily than a tyrosine or tryptophan
residue at position 19. Using the previously reported protocol for
the purification of recombinantly expressed MS2 capsids,4f
60À90 mg of assembled protein could be obtained for each liter
of cultured media. Although modestly lower than those obtained
for wild-type MS2 (∼100 mg/L), these yields were significantly
higher than those obtained when a non-natural residue is
introduced to the same position (10À20 mg/L).
Figure 4. Oxidative coupling of PEG polymers to aromatic residues on
protein substrates. (a) Six MS2 viral capsids were generated with
tryptophan and tyrosine residues on the exterior surface of each
monomer. Two of these capsids were also expressed with a solvent-
accessible cysteine on the interior surface. (b) PEG-substituted pheny-
lene diamine (16, 17) and anisidine (18, 19) derivatives were synthe-
sized and exposed to each viral capsid in the presence of CAN. Reaction
conditions: 1.5 mM CAN, 500 μM PEG, 25 μM protein, 50 mM
bis(TRIS) buffer, pH 6, RT, 1 h. (c) Following capsid disassembly, the
extent of modification for each reaction was determined using SDS-
PAGE with Coomassie staining. (d) Native tyrosine residues (shown in
green) were targeted on chymotrypsinogen under analogous conditions.
(e) The extent of modification for chymotrypsinogen was determined
using SDS-PAGE.
T15W (10), T19W (11), T15Y (12), and T19Y (13) MS2
capsids with high yields and short reaction times (Figure 4c).
Wild-type MS2 was not modified under identical conditions,
confirming that these modifications were dependent on the
presence of the solvent-accessible, exterior tyrosine or trypto-
phan residues. Similarly, treatment of the MS2 mutants with the
PEG-substituted anisidines resulted only in appreciable modification
In order to apply our methodology toward the synthesis of
useful bioconjugates and to evaluate the site-selectivity and bio-
compatibility of our modification strategy, we synthesized a series
of modified PEGs (Figure 4b). In the presence of CAN, the PEG-
substituted phenylene diamines led to the single modification of
16973
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976

Journal of the American Chemical Society
ARTICLE
Figure 5. Double modification of MS2 viral capsids. (a) The cysteine
residues located on the interior surface of capsids 14 and 15 were
modified with Oregon Green maleimide 488 (20). (b) Fluorescently
labeled proteins 21 and 22 were exposed to PEG-substituted anisidine
derivatives (18, 19) under reaction conditions analogous to those shown
in Figure 4. Following capsid disassembly, the extent of modification for
each reaction was determined using SDS-PAGE and visualization with
fluorescent imaging and Coomassie staining.
of T15Y (12) and T19Y (13) MS2. Little to no modification of
wild-type, T15W (10) or T19W (11) MS2 was observed. As was
the case above, the selective modification of the tyrosine mutants
with 18 and 19 occurred with high yields, short reaction times,
and low concentrations of PEG coupling partners (500 μM).
Additionally, analysis by size exclusion chromatography indi-
cated that the capsids remained fully assembled during the
modification process. Up to 77% conversion was observed (on
the basis of densitometry analysis of the Coomassie-stained SDS
PAGE gel), corresponding to the attachment of 138 polymer
chains to each capsid.
Both the PEG-substituted phenylene diamine and anisidine
reagents were also used to modify chymotrypsinogen, a protein
with native, solvent-accessible tyrosine residues (Figure 4d).
Analogous conditions resulted in high yields of modifications
with all PEG substrates (Figure 4e), demonstrating the applic-
ability of our modification strategy to a different protein sub-
strate. In order to confirm the tyrosine selectivity of the reaction,
chymotrypsinogen was reacted with small-molecule reagents 1
and 2. The resulting protein was subjected to trypsin digestion
and analysis by LCÀMS/MS. Only peptides fragments contain-
ing tyrosines were found to contain modifications (Figures S22
and S23 in the SI).
Figure 6. Oxidative coupling of a cell-targeting RGD peptide to
tyrosine residues on the exterior surface of MS2 viral capsids. (a) An
RGD-substituted anisidine derivative (23) was synthesized and exposed
to each viral capsid in the presence of CAN. Reaction conditions:
1.25 mM CAN, 125 μM 23, 25 μM protein, 50 mM bis(TRIS) buffer,
pH 6, 1 h. (b, c) Following capsid disassembly, the extent of modification
for each reaction was determined using SDS-PAGE and MALDI-TOF
MS. (d, e) Fluorescently labeled MS2 was found to undergo modifica-
tion with the RGD derivative under similar conditions. The extent of
modification for each reaction was determined using SDS-PAGE and
visualization with fluorescent imaging and Coomassie staining.
To explore the utility of our reaction for the double modification
of proteins and, in particular, to investigate the compatibility of our
reaction with cysteine alkylation, MS2 double mutants T15Y N87C
(14) and T19Y N87C (15) were generated (Figure 4a). An
introduced cysteine at position 87 has been found to be very
reactive in previous work, and the internal surface of N87C MS2
capsids has been successfully modified with 100À180 copies of
several different maleimide substrates.4d,31À34 Although the
oxidation of unprotected solvent-accessible cysteines and methio-
nineshas been observed in the presence of CAN (resulting in +16
amu adducts), we expected that the alkylation of the cysteines
before exposure to the oxidative conditions could allow for the
dual modification of the capsid. This would avoid oxidative deac-
tivation of these introduced cysteine residues toward further
modification with alkylation reagents. New cysteine residues were
therefore introduced on the interior surface of capsids 12 and 13
at position 87, and the double mutants (14 and 15, respectively)
were isolated with similar yields. To label the proteins with
fluorescent dyes, the introduced cysteines in these capsids were
modified using Oregon Green 488 maleimide (20, Figure 5a).
MALDI-TOF MS of the modified proteins (21 and 22) showed
near-quantitative conversion to a single product (Figure S21 in
the SI).
The treatment of fluorescently labeled proteins 21 (T15Y
N87C) and 22 (T19Y N87C) with the PEG-substituted anisi-
dine derivatives in the presence of CAN resulted in levels of
modification that were comparable to the those observed pre-
viously. In these cases, the cysteine modifications appeared to be
unperturbed by the oxidative conditions, as the PEG-modified
proteins retained their fluorescence (Figure 5b). Reversing the
reaction order for the dual modification of these proteins was also
found to result in the differential modification of the tyrosine and
cysteine groups, but lower yields of both reactions were observed.
Presumably this was a result of the consumption of CAN by the
oxidation of cysteines during the oxidative coupling step, and
the resistance of any oxidized cysteines to modification with
alkylation reagents. Therefore, as expected, in order to use this
chemistry in combination with cysteine modification, it is best
16974
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976

Journal of the American Chemical Society
ARTICLE
to alkylate the thiol group before effecting the oxidative coupling
method.
’ ACKNOWLEDGMENT
These studies were generously supported by the DOD Breast
Cancer Research Program (BC061995). K.L.S. was supported by
a Gerald K. Branch Fellowship in Chemistry. A.C.O. was
supported by a graduate research fellowship from the NSF and
the UC Berkeley Chemical Biology Graduate Program (Training
Grant 1 T32 GMO66698). K.L.S. thanks Tessa Chu for assis-
tance with protein expression and purification and Tony Iavarone
at the UC Berkeley QB3/Chemistry Mass Spectrometry Facility
for assistance with mass spectrometry analysis.
To install potential cancer targeting groups on the surface of
MS2 capsids, small cell-targeting peptides were coupled to the
introduced tyrosine residues using the oxidative coupling strat-
egy. Cyclic RGD is a widely used peptide for the binding of
integrin αvβ3.35À37 A five-amino acid RGD-substituted anisdine
(23, Figure 6a) was synthesized and was shown to couple to
T15Y (12) and T19Y (13) MS2 with high yields and selectivity
(Figure 6b,c). Little to no modification was observed for wild-
type MS2, indicating that the solvent-accessible, exterior tyrosine
was needed for the reaction to occur. Exposure of fluorescently
labeled capsids 21 and 22 to the RGD-substituted anisidine
using analogous reaction conditions also resulted in modifica-
tion, generating assemblies that were functionalized with both
an internal imaging group and an exterior targeting group
(Figure 6d,e). Although the application of this methodology
for the attachment of peptides to proteins will be limited to
sequences without tyrosine residues, the ability of this unprece-
dented reaction to reach appreciable levels of conversion despite
the significant amount of steric hindrance suggests that it will find
use in many circumstances.
’ REFERENCES
(1) (a) Hermanson, G. T. Bioconjugate Techniques, 2nd ed.; Aca-
demic Press: San Diego, 2008. (b) Tilley, S. D.; Joshi, N. S.; Francis,
M. B. The Chemistry And Chemical Reactivity Of Proteins. In The Wiley
Encyclopedia of Chemical Biology; Begley, T., Ed; Wiley-VCH: Wein-
heim, 2008.
(2) Doolittle, R.F. R. Redundancies in Protein Sequence. In Predic-
tion of Protein Structures and the Principles of Protein Conformation;
Fasman, G. D., Ed; Plenum Press: New York, 1989.
(3) Examples of applications that require modifications in two or
more distinct locations include F€orster resonance energy transfer
(FRET) experiments,3a folding studies,3b or the dual modification of
proteins for the targeted delivery of drugs or imaging agents.3c,d(a) Joo,
C.; Balci, H.; Ishitsuka, Y.; Buranachai, C.; Ha, T. Annu. Rev. Biochem.
2008, 77, 51–76. (b) Cecconi, C.; Shank, E. A.; Bustamante, C.;
Marqusee, S. Science 2005, 309, 2057–2060. (c) Banerjee, D.; Liu,
A. P.; Voss, N. R.; Schmid, S. L.; Finn, M. G. ChemBioChem 2010, 11,
1273–1279. (d) Stephanopoulos, N.; Tong, G. J.; Hsiao, S. C.; Francis,
M. B. ACS Nano 2010, 4, 6014–6020.
(4) Examples of adding new functionality to artificial amino acids
have involved methods targeting ketones,4a azides,4b,c,d alkynes,4b,d,e and
anilines.4f,g(a) Cornish, V. W.; Hahn, K. M.; Schultz, P. G. J. Am. Chem.
Soc. 1996, 118, 8150–8151. (b) Deiters, A.; Cropp, T. A.; Mukherji, M.;
Chin, J. W.; Anderson, J. C.; Schultz, P. G. J. Am. Chem. Soc. 2003,
125, 11782–11783. (c) Kiick, K. L.; Saxon, E.; Tirrell, D. A.; Bertozzi,
C. R. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 19–24. (d) Strable, E.;
Prasuhn, D. E.; Udit, A. K.; Brown, S.; Link, A. J.; Ngo, J. T.; Lander, G.;
Quispe, J.; Potter, C. S.; Carragher, B.; Tirrell, D. A.; Finn, M. G.
Bioconjugate Chem. 2008, 19, 866–875. (e) van Hest, J. C. M.; Kiick,
K. L.; Tirrell, D. A. J. Am. Chem. Soc. 2000, 122, 1282–1288. (f) Carrico,
Z. M.; Romanini, D. W.; Mehl, R. A.; Francis, M. B. Chem. Commun.
2008, 1205–1207. (g) Mehl, R. A.; Anderson, J. C.; Santoro, S. W.;
Wang, L.; Martin, A. B.; King, D. S.; Horn, D. M.; Schultz, P. G. J. Am.
Chem. Soc. 2003, 125, 935–939.
(5) Antos, J. M.; Francis, M. B. J. Am. Chem. Soc. 2004, 126, 10256–
10257.
(6) Hooker, J. M.; Kovacs, E. W.; Francis, M. B. J. Am. Chem. Soc.
2004, 126, 3718–3719.
(7) Joshi, N. S.; Whitaker, L. R.; Francis, M. B. J. Am. Chem. Soc.
2004, 126, 15942–15943.
(8) Tilley, S. D.; Francis, M. B. J. Am. Chem. Soc. 2006, 128, 1080–1081.
(9) Antos, J. M.; McFarland, J. M.; Iavarone, A. T.; Francis, M. B.
J. Am. Chem. Soc. 2009, 131, 6301–6308.
(10) Popp, B. V.; Ball, Z. T. J. Am. Chem. Soc. 2010, 132, 6660–6662.
(11) Foettinger, A.; Leitner, A.; Lindner, W. J. Proteome Res. 2007,
6, 3827–3834.
(12) Foettinger, A.; Melmer, M.; Leitner, A.; Lindner, W. Bioconju-
gate Chem. 2007, 18, 1678–1683.
’ CONCLUSION
This work introduces several new oxidative coupling methods
for the chemoselective modification of aromatic amino acids in
proteins. These reactions have not been reported previously, and
thus add to the important and growing list of useful bioconjuga-
tion techniques. The ability of these reactions to target natural
amino acids provides a useful complement to strategies that
require the use of artificial amino acids. Similar levels of chemo-
selectivity and yield can be achieved, provided that there are few
surface-accessible tyrosines and tryptophans in the protein
substrate. Although not the focus of this work, the new
tryptophan reactivity provides very important design leads for
future reactions, as there are few existing bioconjugation
methods for this residue.5,9À13 In comparison to other tyrosine
modification strategies, the anisidine coupling reaction pro-
ceeds with excellent chemoselectivity and moderate to low
concentrations of simple reagents. It avoids the elevated pH
required for diazonium coupling reactions and does not modify
histidine residues.1 Futhermore, it generally reaches higher levels
of conversion in shorter time periods than previously reported
Mannich-type couplings7 and π-allylpalladium alkylations.8 We
therefore envision these new oxidative coupling reactions to be
useful in many situations that require bioconjugation reactions to
be run on large scale, such as the creation of new protein-based
materials.
’ ASSOCIATED CONTENT
S
Supporting Information. Full experimental details, addi-
b
tional characterization spectra, protein digest information, and
current mechanistic hypotheses for these transformations. This
material is available free of charge via the Internet at http://pubs.
acs.org.
(13) Sturm, M.; Leitner, A.; Lindner, W. Bioconjugate Chem. 2011,
22, 211–217.
(14) Ban, H.; Gavrilyuk, J.; Barbas, C. F. J. Am. Chem. Soc. 2010,
132, 1523–1525.
’ AUTHOR INFORMATION
(15) Davies, M. J. Biochim. Biophys. Acta-Proteins Proteomics 2005,
1703, 93–109.
(16) Davies, M. J. Biochem. Biophys. Res. Commun. 2003, 305, 761–770.
Corresponding Author
francis@cchem.berkeley.edu
16975
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976

Journal of the American Chemical Society
ARTICLE
(17) Dean, R. T.; Fu, S.; Stocker, R.; Davies, M. J. Biochem. J. 1997,
324, 1–18.
(18) Ahlbrink, R.; Haumann, M.; Cherepanov, D.; B€ogershausen,
O.; Mulkidjanian, A.; Junge, W. Biochemistry 1998, 37, 1131–1142.
(19) Hoganson, C. W.; Babcock, G. T. Science 1997, 277, 1953–1956.
(20) Renger, G. Biochim. Biophys. Acta-Bioenergetics 2004, 1655,
195–204.
(21) Stubbe, J.; van Der Donk, W. A. Chem. Rev. 1998, 98, 705–762.
(22) Brown, K. C.; Kodadek, T. Met. Ions Biol. Syst. 2001, 38,
351–384.
(23) Fancy, D. A.; Denison, C.; Kim, K.; Xie, Y.; Holdeman, T.;
Amini, F.; Kodadek, T. Chem. Biol. 2000, 7, 697–708.
(24) Fancy, D. A.; Kodadek, T. Proc. Natl. Acad. Sci. U.S.A. 1999,
96, 6020–6024.
(25) We have previously shown that N,N-dialkylphenylene diamines
and anisidine derivatives couple to introduced aniline groups on proteins
in the presence of sodium periodate.25a,b These conditions lead to
minimal if any modification of native residues. See Figure S4 in the
Supporting Information for MALDI-TOF MS spectra of melittin and
angiotensin after subjection to analogous coupling conditions using
sodium periodate as the oxidant. (a) Hooker, J. M.; Esser-Kahn, A. P.;
Francis, M. B. J. Am. Chem. Soc. 2006,128, 15558–15559. (b) Stephanopoulos,
N.; Carrico, Z. M.; Francis, M. B. Angew. Chem., Int. Ed. 2009, 48, 9498–9502.
See also reference 4f.
(26) Although the use of cerium (IV) ammonium nitrate has been
reported for the deprotection of p-methoxy benzyl ethers in aqueous
solutions,26a,b we found our anisidine derivatives to be stable in the
presence of CAN over a period of 5 h. (a) Classon, B.; Garegg, P. J.;
Samuelsson, B. Acta Chem. Scand. B 1984, B38, 419–422. (b) Johansson,
R.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1 1984, 2371–2374.
(27) For a general reference on cerium chemistry, see: Binnemans,
K. Applications of Tetravalent Cerium Compounds. In Handbook on the
Physics and Chemistry of Rare Earths; Gschneidner, K. A., Jr., B€unzli, J.-C.
G., Pecharsky, V. K., Eds; Elsevier B.V.: Amsterdam, 2006.
(28) Phosphate buffer was not screened; the addition of cerium(IV)
ammonium nitrate to phosphate buffer solutions results in the immedi-
ate precipitation of the cerium species. As result, the oxidant is
deactivated, and the coupling reactions are unable to proceed. For the
successful application of this methodology, it is therefore necessary to
remove any phosphate from the reaction mixture.
(29) Valegard, K.; Liljas, L.; Fridborg, K.; Unge, T. Nature 1990,
345, 36–41.
(30) Peabody, D. S. J. Nanobiotechnol. 2003, 1, 5.
(31) Meldrum, T.; Seim, K. L.; Bajaj, V. S.; Palaniappan, K. K.; Wu,
W.; Francis, M. B.; Wemmer, D. E.; Pines, A. J. Am. Chem. Soc. 2010,
132, 5936–5937.
(32) Stephanopoulos, N.; Carrico, Z. M.; Francis, M. B. Angew.
Chem., Int. Ed. 2009, 48, 9498–9502.
(33) Tong, G. J.; Hsiao, S. C.; Carrico, Z. M.; Francis, M. B. J. Am.
Chem. Soc. 2009, 131, 11174–11178.
(34) Wu, W.; Hsiao, S. C.; Carrico, Z. M.; Francis, M. B. Angew.
Chem., Int. Ed. 2009, 48, 9493–9497.
(35) Pierschbacher, M. D.; Ruoslahti, E. Nature 1984, 309, 30–33.
(36) Auzzas, L.; Zanardi, F.; Battistini, L.; Burreddu, P.; Carta, P.;
Rassu, G.; Curti, C.; Casiraghi, G. Curr. Med. Chem. 2010, 17, 1255–1259.
(37) Liu, S. Mol. Pharmaceutics 2006, 3, 472–487.
16976
dx.doi.org/10.1021/ja206324q |J. Am. Chem. Soc. 2011, 133, 16970–16976