REVIEW

Chemistry in living systems

© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology

Jennifer A Prescher1 & Carolyn R Bertozzi1–4

Dissecting complex cellular processes requires the ability to track biomolecules as they function within their native habitat.
Although genetically encoded tags such as GFP are widely used to monitor discrete proteins, they can cause significant
perturbations to a protein’s structure and have no direct extension to other classes of biomolecules such as glycans, lipids, nucleic
acids and secondary metabolites. In recent years, an alternative tool for tagging biomolecules has emerged from the chemical
biology community—the bioorthogonal chemical reporter. In a prototypical experiment, a unique chemical motif, often as small
as a single functional group, is incorporated into the target biomolecule using the cell’s own biosynthetic machinery. The chemical
reporter is then covalently modified in a highly selective fashion with an exogenously delivered probe. This review highlights the
development of bioorthogonal chemical reporters and reactions and their application in living systems.

Living systems are composed of networks of interacting biopolymers,
ions and metabolites. These components drive a complex array of cellu-
lar processes, many of which cannot be observed when the biomolecules
are examined in their purified, isolated forms. Accordingly, researchers
have begun moving beyond the artificial confines of test tubes to study
biological processes in the context of living cells and whole organisms.
This endeavor requires the ability to track molecules within their native
environs. Few biomolecules are naturally endowed with features that
permit their direct detection in complex milieus. Thus, several methods
have been developed to equip cellular components with reporter tags
for visualization and isolation from biological samples.
The most popular tagging strategy for cellular imaging involves the
use of the green fluorescent protein (GFP) and its related variants1–3.
The fusion of these fluorescent probes to a target protein enables
visualization by fluorescence microscopy and quantification by flow
cytometry. Because they are genetically encoded and require no auxil-
iary cofactors, GFP tags can be used to analyze protein expression and
localization in living cells and whole organisms4,5. Almost every cel-
lular process has been interrogated using fluorescent protein fusions,
including glycoprotein transport in the secretory pathway6 and tran-
scription in the nucleus7. Furthermore, a collection of GFP-like tags is
now available with emission wavelengths that span virtually the entire
visible spectrum8–10.
Although fluorescent protein fusions are undoubtedly the most pow-
erful general tools for imaging proteins within living systems, they are
not without limitations. These relatively large proteins can be a signifi-
cant structural perturbation and may therefore influence the expression,
localization or function of the protein to which they are attached. Also,
fluorescent protein fusions can be visualized only by optical methods,
without an obvious extension to other imaging modalities. Finally, GFP
Departments of 1Chemistry and 2Molecular and Cell Biology and 3Howard
Hughes Medical Institute, University of California, Berkeley, California
94720, USA. 4Materials Sciences Division, Lawrence Berkeley National
Laboratory, Berkeley, California 94720, USA. Correspondence should be
addressed to C.R.B. (crb@berkeley.edu).
variants cannot be applied to visualization of glycans, lipids, nucleic
acids or the thousands of small organic metabolites amassed within
cells (Fig. 1). Non-proteinaceous materials comprise a significant frac-
tion of cellular biomass11, and the ability to image these species would
augment our understanding of cellular biochemistry. Glycans, lipids
and inorganic ions are also involved in modulating protein activity
by post-translational modification12. Therefore, methods to visualize
both proteins and their modifiers would contribute to a more holistic
understanding of the proteome.
Antibody conjugates have been widely used to track biomolecules in
living cells and whole organisms13. They can be generated with specific-
ity for virtually any epitope and are therefore, in principle, applicable
to imaging a wide range of biomolecules. However, the large size and
physical properties of these reagents hinder their access to antigens
within cells and outside of the vasculature in living animals14,15.
In general, small molecules have better access to intracellular and
extravascular compartments. Their use as imaging agents requires a
means to selectively target the small probe to a desired biomolecule.
Nucleophilic functionality occurs in most types of biopolymers, per-
mitting facile derivatization with biotin, fluorophores and numerous
other small-molecule reporters. Established bioconjugation protocols
have made these operations trivial for purified biopolymers in vitro16.
However, the site-specific chemical modification of biomolecules
within their native settings remains a formidable challenge.
In recent years, an alternative strategy for tagging biomolecules has
emerged that blends the simplicity of genetically encoded tags with the
specificity of antibody labeling and the versatility of small-molecule
probes. This approach involves the incorporation of unique chemical
functionality—a bioorthogonal chemical reporter—into a target bio-
molecule using the cell’s own biosynthetic machinery. Bioorthogonal
chemical reporters are non-native, non-perturbing chemical handles
that can be modified in living systems through highly selective reactions
with exogenously delivered probes. This two-step labeling process can
be used to outfit a target biomolecule for detection or isolation, depend-
ing on the nature of the probe. Proteins17–20, glycans21–24 and lipids25
have all been fashioned with an assortment of chemical reporters in
living cells and subsequently ligated with reactive probes. Most recently,

NATURE CHEMICAL BIOLOGY VOLUME 1 NUMBER 1 JUNE 2005 13

 REVIEW
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
Figure 1 Composition of a typical mammalian cell11. Although proteins
comprise the largest fraction of a cell’s dry mass, it is estimated that more
than half are modified with glycans, lipids or other metabolites113. Methods
for visualizing both proteins and non-proteinaceous biomolecules would
enhance our understanding of living systems.
the chemical reporter strategy has been applied to monitoring enzyme
activities26–29 and tagging cell surface glycans in whole organisms30.
The breadth of these examples underscores the impact of bioorthogonal
chemical reporters in expanding the repertoire of biomolecules that can
be visualized in living systems.
Here we summarize the development of bioorthogonal chemi-
cal reporters and their applications in biology. First, we provide an
overview of existing chemical reporters and bioorthogonal reactions.
Second, we discuss the applications of these chemistries to monitoring
biomolecules and enzyme activities in cellular systems. Third, we high-
light the translation of one chemical reporter system from cell-based
studies to living animals. Last, we outline future challenges in the field,
from the perspective of both chemists and biologists.
Design of chemical reporters and bioorthogonal reactions
The bioorthogonal chemical reporter strategy involves the incorpora-
tion of unique functionality into targets of interest, followed by chemi-
cal labeling with a small-molecule probe (Fig. 2). Ideally, the chemical
reporter (blue circle, Fig. 2) should be integrated into the target scaffold
without significant structural perturbation. This is accomplished by
appending the reporter to substrates that can be used by the cell’s own
metabolic machinery. For example, amino acids bearing bioorthogo-
nal functional groups can be accepted by the translational machin-
ery of a cell and incorporated into proteins. Similarly, functionalized
monosaccharides can be introduced into cell surface glycans by means
of promiscuous enzymes in the biosynthetic pathways of these bio-
H2O
Figure 2 The bioorthogonal chemical reporter
strategy. A chemical reporter (blue circle) linked
to a substrate (light green box) is introduced Target
into a target biomolecule through cellular
incorporation
metabolism. In a second step, the reporter is
covalently tagged with an exogenously delivered O2
probe (blue arc). Both the chemical reporter and
exogenous probe must avoid side reactions with
nontarget biomolecules (gray shapes). = bioorthogonal chemical reporter
14
polymers. Regardless of the route exploited, each enzyme involved in
the installation process must tolerate the unnatural motif. For this rea-
son, typical biophysical probes, such as fluorescein, cannot be used as
direct modifications to metabolic substrates (that is, amino acids, lipids
or sugars) as their relatively large size would interfere with enzymatic
transformations. A small functional group is more likely to be tolerated
by metabolic enzymes. Thus, to date, bioorthogonal chemical report-
ers have been non-native combinations of endogenous functionality
(as discussed below) or small, abiotic functional groups that can slip
through existing biosynthetic pathways.
Once installed in a target biomolecule, the chemical reporter must
be reacted with a probe bearing a complementary chemical moiety
(blue arc, Fig. 2). The requirements for the covalent reaction between
the two components are quite stringent. The reporter and its part-
ner must be mutually reactive in a physiological environment (37 °C,
pH 6–8) and, at the same time, remain inert to the surrounding biologi-
cal milieu. Ideally, the reactants should function similar to an antibody-
antigen duo, reacting rapidly with one another, unaided by auxiliary
reagents, to form a stable adduct with innocuous (or no) byproducts.
Considering the abundance of nucleophiles, reducing agents and other
functionality present in cells, the choice of suitable components for the
chemical transformation is far from obvious. For instance, amines and
isothiocyanates, thiols and maleimides, and other coupling partners
typically used for bioconjugation must be avoided to prevent labeling
of irrelevant targets. In addition, the chemical reporter and its comple-
mentary probe must possess adequate metabolic stability and bioavail-
ability for use in cells or organisms.
Existing bioorthogonal chemical reporters
So far, only a handful of chemical motifs are known to possess the re-
quisite qualities of biocompatibility and selective reactivity to function
as bioorthogonal chemical reporters in living cells. This elite group com-
prises peptide sequences that can be ligated with small-molecule imaging
probes18,31, cell surface electrophiles that can be tagged with hydrazide
and aminooxy derivatives19,22, azides that can be selectively modified
with phosphines32 or activated alkynes33,34, and terminal alkynes that
can be ligated with azides (Table 1)29. The sections that follow introduce
each of these chemical reporters and summarize their advantages and
disadvantages in tagging biomolecules in cellular systems.
Bioorthogonal peptide sequences. No single proteogenic amino acid
side chain can function as a unique chemical moiety for target-specific
tagging. However, Tsien and coworkers have demonstrated that unique
combinations of side chains can create new functionality that satis-
fies the criteria of a bioorthogonal chemical reporter. They designed
a short peptide sequence containing a tetracysteine motif (CCXXCC,
where XX are virtually any two amino acids, but optimally proline and
glycine) that reacts selectively with biarsenicals18,31. The hexapeptide
chemical reporter can be fused to target proteins at the genetic level and
covalently labeled in living cells with membrane-permeant biarsenical
H2O H2O
Metabolic Chemical
labeling
O2 O2
VOLUME 1 NUMBER 1 JUNE 2005 NATURE CHEMICAL BIOLOGY
 REVIEW

Table 1 Chemical reporters and bioorthogonal reactions used in

living systems.

Chemical Reactive partner Ligation product Targeta

reporter (R´ = probe) (R)
R
R S S S S
As As X = N, ReAsH
HO O O
S S S
S Protein18,31
HS
HS As As
SH X
SH HO O O
CO2
X = , FlAsH
Tetracysteine motif X
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology

H
O N R'
N
O H 2N N R' O
H R R"(H) Protein19,20
R R"(H) O R' Glycan22
N
H 2 N O R'
Ketone/aldehyde
R R"(H)

Staudinger ligation O
O Figure 3 Bioorthogonal chemical reporters and cellular imaging. HeLa cells
RHN
H 3CO
R'
expressing tetracysteine-fused connexin were treated with FlAsH (green),
R' Ph 2P incubated in medium for 4 hours, then treated with ReAsH (red) and
Ph 2 P O O Protein17,26
O imaged. This two-color pulse-chase labeling experiment demonstrated that
R N3 ‘Click’ chemistry N N Glycan30,34 newly synthesized connexin is incorporated at the outer edges of existing gap
Azide R' , Cu(I), ligand R N R' Lipid25
junctions (indicated by white arrows)37. Figure reproduced from ref. 37 by
Strain-promoted cycloaddition
permission of the American Association for the Advancement of Science.
N N
N
R' R
R'

‘Click’ chemistry tag is a minimal structural perturbation relative to fluorescent proteins.

N N
R
N 3 R' , Cu(I), ligand
R
Terminal alkyne
a
Only literature examples provided. Other biomolecules could potentially be labeled in
a similar manner.
dyes, such as the fluorescein derivative FlAsH and the resorufin deriva-
tive ReAsH (Table 1). The ethanedithiol substituents of these reagents
prevent the labeling of biomolecules bearing isolated cysteine residues.
Furthermore, the biarsenical probes are only weakly fluorescent when
free in solution and undergo a marked increase in fluorescence when
bound to the target sequence. Target singularity is ensured by the rarity
of the hexapeptide motif among endogenous proteins.
The tetracysteine reporter group has been used to image a variety of
proteins, including some whose distribution was known to be perturbed
by GFP labeling35,36. In several studies, combinations of FlAsH and ReAsH
were used to observe the real-time assembly, trafficking and degradation
of the protein targets (Fig. 3)37,38. ReAsH also doubles as a photosensitizer,
generating singlet oxygen to selectively inactivate proteins to which it is
fused39 or to produce contrast stains for electron microscopy37.
The tetracysteine-biarsenical method has also inspired the deve-
lopment of several complementary approaches for attaching small
molecules to proteins40–43. Many of these strategies involve enzymatic
reactions or ligand-receptor binding. For example, proteins can be
labeled by fusion to an enzyme (for example, human O6-alkylguanine
transferase44) or receptor (for example, FKBP12(F36V)45, dihydrofo-
late reductase46,47) that is capable of binding functionalized probes.
Additionally, proteins can be labeled by fusion to peptide sequences
that bind small-molecule reagents. These include histidine-rich pep-
tides recognized by functionalized Ni-NTA probes48, peptide aptamers
engineered to bind the fluorophore Texas Red49 and acidic peptides
that can bind luminescent lanthanides50. Muir and coworkers have also
reported the use of trans-splicing inteins for tagging proteins in living
cells51. Further optimization of all these labeling methodologies may
permit their more widespread application in biological systems.
In summary, the tetracysteine-biarsenical system affords a powerful
alternative to GFP tagging for protein visualization. The hexapeptide
N R' Protein29 Still, as a modification to metabolic substrates such as amino acids and
monosaccharides, the hexapeptide tag is unlikely to be tolerated by
biosynthetic enzymes. Thus, metabolic labeling of biopolymers other
than proteins requires an alternative—and even smaller—chemical
reporter. As described below, carefully chosen, simple functional groups
can fulfill this purpose.
Ketones and aldehydes. Comprising only a handful of atoms, ketones and
aldehydes are bioorthogonal chemical reporters that can tag not only pro-
teins, but also glycans and other secondary metabolites (Table 1). These
mild electrophiles are attractive choices for modifying biomolecules as
they are readily introduced into diverse scaffolds, absent from endogenous
biopolymers and essentially inert to the reactive moieties normally found
in proteins, lipids and other macromolecules. Although these carbonyl
compounds can form reversible Schiff bases with primary amines such
as lysine side chains, the equilibrium in water favors the carbonyl. By
contrast, the stabilized Schiff bases with hydrazide and aminooxy groups
(hydrazones and oximes, respectively) are favored in water and are quite
stable under physiological conditions52.
Rideout and coworkers recognized the potential use of ketones and
aldehydes for chemoselective drug assembly in the presence of living
cells53–55. They reported that decanal and octyl aminoguanidine—both
independently harmless to cells—react selectively to form a hydrazone-
linked detergent capable of lysing cultured erythrocytes. This same
strategy was used to generate inhibitors of protein kinase C from the
in situ assembly of aldehyde and hydrazide precursors56. As described
in more detail later, this transformation has been used to chemically
modify mammalian cell surfaces19,22,23,57,58. More recently, Sadamoto
and coworkers introduced ketones into bacterial cell walls and labeled
the reporters with a hydrazide-based fluorophore59.
Although suitable for chemical modifications in the presence of
cultured cells, ketone (and aldehyde) condensations are somewhat
limited in the context of living organisms. The pH optimum of these
reactions is 5–6, values that cannot be achieved in most tissues in vivo.
Additionally, ketones and aldehydes are not truly bioorthogonal in
more complex physiological settings. Keto and aldehydic metabolites

NATURE CHEMICAL BIOLOGY VOLUME 1 NUMBER 1 JUNE 2005 15

 REVIEW
O N2 O Staudinger ligation. In 1919, Hermann Staudinger reported that azides
OCH3 OCH3
+ N3 R'
R P R P N R'
Ph Ph
Ph Ph
Phosphine Azide Aza-ylide
CH3 OH
O
O intramolecular trap into the phosphine (Fig. 4)32. Now known as the
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
N R'
H H2O N R'
R P O
Ph R P
Ph
Ph Ph
Ligation product
Figure 4 The Staudinger ligation. A triarylphosphine and an azide first react
to form an aza-ylide intermediate. The nucleophilic nitrogen atom is trapped
in an intramolecular fashion, and the cyclized intermediate hydrolyzes in
water to form a stable amide-linked product. In some cases, aryl azides
(R’ = aryl) may react with phosphines to initially form O-alkyl imidates114.
are abundant within cells and in biological fluids in the form of free
sugars, pyruvate, oxaloacetate and various cofactors (such as pyridoxal
phosphate). Therefore, aldehydes and ketones are best used in environs
devoid of carbonyl electrophiles (namely, on cell surfaces or in the
extracellular environment) and should be considered ‘biorestricted’
chemical reporters.
Azides. In contrast to aldehydes and ketones, azides are viable chemi-
cal reporters for labeling all classes of biomolecules in any biological
locale (Table 1). This versatile functional group is abiotic in animals
and absent from nearly all naturally occurring species. (Only one
naturally occurring azido metabolite has been reported to date, iso-
lated from unialgal cultures.)60 Azides do not react appreciably with
water and are resistant to oxidation. Additionally, azides are mild
electrophiles; but unlike aldehydes, they do not react with amines or
the other ‘hard’ nucleophiles that are abundant in biological systems.
Rather, they require ‘soft’ nucleophiles for reaction. Azides are there-
fore susceptible to reduction by free thiols, including the ubiquitous
cellular reductant, glutathione. However, reactions between mono-
thiols and alkyl azides typically require vigorous heating (100 °C for
several hours) or auxiliary catalysts60,61.
Despite its exquisite bioorthogonality, the azide has only recently
been used as a chemical reporter in living systems. This may be due
to perceptions of the azide as unstable, toxic or both. Azides are prone
to decomposition at elevated temperatures, but they are quite stable
at physiological temperatures60. Whereas aryl azides are well-known
photocrosslinkers, alkyl azides do not photodecompose in the presence
of ambient light. Finally, although azide anion (for example, in the form
of NaN3) is a widely used cytotoxin, organic azides have no intrinsic
toxicity. Indeed, organic azides are components of clinically approved
drugs such as AZT60.
Although kinetically stable, azides are predisposed to unique modes
of reactivity owing to their large intrinsic energy content. This fea-
ture has been exploited for the development of bioorthogonal reac-
tions, including the Staudinger ligation of azides with functionalized
phosphines and the [3+2] cycloaddition of azides with activated
alkynes. These reactions can be used for the selective labeling of azide-
functionalized biomolecules.
16
react with triphenylphosphines (soft nucleophiles) under mild condi-
tions to produce aza-ylide intermediates62. These intermediates can
be subsequently hydrolyzed in water or trapped by myriad electro-
philes to provide a pair of products: an amine and the corresponding
phosphine oxide63. The bioorthogonal nature of this transformation
suggested potential applications of the azide as a chemical reporter,
provided a covalent link could be forged between the two reactants.
We modified the classic Staudinger reaction by introduction of an
Staudinger ligation, this transformation ultimately produces a covalent
link between one nitrogen atom of the azide and the triarylphosphine
scaffold. The Staudinger ligation can be used to covalently attach probes
to azide-bearing biomolecules. Like the azide, phosphines do not react
appreciably with biological functional groups and are therefore also
bioorthogonal. Additionally, the reaction proceeds readily at pH 7 with
no apparent toxic effects. Oxidation of the phosphine by air or meta-
bolic enzymes is the only potentially problematic side reaction that may
diminish the amount of probe that is available in biological systems.
The Staudinger ligation has been used to modify glycans on living
cells32, enrich glycoprotein subtypes from various proteomes64,65 and
impart new functionality to recombinant proteins66. Raines and col-
leagues and researchers in our laboratory developed phosphine reagents
for a related transformation that produces amide-linked products with-
out incorporation of the phosphine oxide into the final adducts67,68.
Although these phosphines have not been used for bioconjugation in
living systems, they have been used to immobilize small molecules69
and proteins70 on glass slides. These and other applications of the
Staudinger ligation have been recently reviewed71.
Copper-catalyzed [3+2] azide-alkyne cycloaddition. In the context of the
Staudinger ligation, the azide serves as an electrophile subject to reac-
tion with soft nucleophiles. Azides are also 1,3-dipoles that can undergo
reactions with dipolarophiles such as activated alkynes72. These π-
systems are both extremely rare and inert in biological systems, further
enhancing the bioorthogonality of the azide along this reaction trajec-
tory. The [3+2] cycloaddition between azides and terminal alkynes to
provide stable triazole adducts was first described by Huisgen more
than four decades ago73. The reaction is thermodynamically favorable
by an impressive 30–35 kcal/mol. Without alkyne activation, however,
the process requires elevated temperatures or pressures that are not
compatible with living systems. How can activation be achieved? One
possibility involves the addition of electron-withdrawing groups (such
as esters) to the alkyne. Unfortunately, the resulting α,β-unsaturated
carbonyl compounds can also act as Michael acceptors for a variety of
biological nucleophiles and are therefore not bioorthogonal.
Another possibility involves the use of a catalyst. Sharpless and cowork-
ers and Meldal and coworkers demonstrated that the rate of cycload-
dition between azides and alkynes can be accelerated ∼106-fold using
catalytic amounts of Cu(I)74,75. This copper-catalyzed reaction, termed
‘click’ chemistry, proceeds readily at physiological temperatures and in
the presence of biological materials to provide 1,4-disubstituted triazoles
with nearly complete regioselectivity (Table 1)33. The copper-mediated
reaction has been used to tag azides installed within virus particles76,
nucleic acids77 and proteins from complex tissue lysates78 with virtually
no background labeling. It should be noted that the same reaction can
be carried out using the alkyne as the chemical reporter (Table 1). Like
the azide, a terminal alkyne consists of a mere three atoms.
The primary advantage of the catalyzed azide-alkyne cycloaddition over
the Staudinger ligation is its faster rate. Based on preliminary studies in our
laboratory, the copper-catalyzed reaction of azides with alkynes proceeds
VOLUME 1 NUMBER 1 JUNE 2005 NATURE CHEMICAL BIOLOGY
 REVIEW

O
O H
a Residue-specific incorporation ‘Keto-biotin’ H
S
H H
N3 CO2H
S

ATP, BirA
Methionine
H2 N CO2 H H2N CO2H
analogs
O c Chemical
N3 reporter tags
Phenylalanine FGE
O
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology

analogs
H2 N CO 2 H H2N CO2H
H

N3 O

O
H2 N CO2H H2 N CO 2 H H2 N CO2 H

b Site-specific incorporation

Figure 5 Methods for introducing chemical reporters into proteins. (a) Unnatural amino acids bearing ketones, azides and alkynes can be incorporated into
target proteins in a residue-specific manner using auxotrophic strains of E. coli. (b) Amino acids with bioorthogonal side chains can be installed into proteins
in a site-specific fashion using nonsense suppression techniques. (c) Chemical reporters can be introduced into short peptide sequences using the cell’s
post-translational machinery. In one example, an analog of biotin (‘keto-biotin’) is attached to a 15-amino-acid consensus sequence (blue box) by E. coli
biotin ligase (BirA). Similarly, formylglycine-generating enzyme (FGE) can convert a cysteine residue within a 13-residue consensus sequence (red box) to
formylglycine. Both of the these electrophiles can be labeled with hydrazide probes.

at least 25 times faster than the reaction of azides with triarylphosphines
in cell lysates. Accordingly, ‘click’ chemistry has been used in situations
that require detection of very small quantities of azide-labeled biomol-
ecules78. The primary disadvantage of the copper-catalyzed cycloaddition
is the cellular toxicity of the metal catalyst79. Although more biofriendly
metal-ligand combinations could potentially be discovered, the reaction
is not ideal for labeling biomolecules in living cells.
Strain-promoted cycloaddition. An alternative means of activating
alkynes for catalyst-free [3+2] cycloaddition with azides involves the
use of ring strain34. Constraining the alkyne within an eight-membered
ring creates ∼18 kcal/mol of strain, much of which is released in the
transition state upon [3+2] cycloaddition with an azide80. As a conse-
quence, cyclooctynes react with azides at room temperature, without
the need for a catalyst81. This strain-promoted cycloaddition has been
used to label biomolecules both in vitro and on cell surfaces without
observable toxic effects34. However, the reaction is limited by its slow
rate. (The second-order rate constant for the reaction of a derivatized
cyclooctyne with benzyl azide in aqueous CD3CN is 0.0012 M–1 s–1,
whereas that for the Staudinger ligation is 0.0025 M–1 s–1; refs. 34,82.)
Preliminary results from our laboratory indicate that the rate of the
strain-promoted cycloaddition can be increased by appending electron-
withdrawing groups to the octyne ring (C.R.B., unpublished data).
Introducing ketones, azides and alkynes into biomolecules
Proteins. Ketones, azides and alkynes are not included in the repertoire
of side chain functional groups found in the 20 proteogenic amino acids.
To exploit their bioorthogonal chemistry for protein labeling requires a
means for de novo introduction of these chemical reporters, typically in
the form of unnatural amino acids (Fig. 5). This can be accomplished
using a cell’s translational machinery in either a residue-specific83 or a
site-specific manner84. As described by Tirrell and coworkers, residue-
specific incorporation of unnatural amino acids into proteins simply
involves replacement of a natural residue with a conservatively modified
analog (Fig. 5a). The translational machinery is sufficiently tolerant of
altered substrates that, in the absence of competing natural substrates,
the modified residue is converted to an aminoacyl tRNA that is subse-
quently used by the ribosome. By this mechanism, unnatural amino
acids bearing bioorthogonal chemical reporters can be introduced into
proteins that are overexpressed in Escherichia coli. To avoid competition
with the endogenous residue, the bacterial strain is rendered auxotrophic
for the natural amino acid. Proteins cannot be overexpressed unless
the cells are supplemented with either that residue or a closely related
unnatural analog. For example, a phenylalanine auxotroph was used to
express proteins in which all phenylalanine residues were replaced with
p-azidophenylalanine or p-acetylphenylalanine (a keto derivative)85,86.
Similarly, a methionine auxotroph was used for production of proteins
that contained homopropargylglycine or azidohomoalanine at sites that
encode for methionine87,88. Notably, Link et al. have extended this work
to the labeling of bacterial cell surfaces17,79. Azido amino acids were
installed in outer membrane protein C (OmpC) of an E. coli methio-
nine auxotroph and the cell surface azides were then ligated with alkyne
probes through both copper(I)-mediated and strain-promoted [3+2]
cycloaddition (D.A. Tirrell, personal communication; ref. 79).
Residue-specific metabolic labeling can produce proteins with mul-
tiple copies of a bioorthogonal functional group, but it has only limited
application in cases where a chemical reporter is desired at a single
position within the protein. As pioneered by Schultz and coworkers,
site-specific insertion of a bioorthogonal amino acid has been achieved
using nonsense suppression techniques (Fig. 5b)84. In this approach, a
mutually selective tRNA and aminoacyl-tRNA synthetase are developed
so that the unnatural amino acid can be uniquely activated by the tRNA
in vivo. The tRNA’s anticodon is engineered to complement a rare stop
codon, which is co-opted to encode the unnatural amino acid in the
corresponding DNA (and intermediate mRNA). Cells transfected with
genes encoding the engineered tRNA, aminoacyl-tRNA synthetase and

NATURE CHEMICAL BIOLOGY VOLUME 1 NUMBER 1 JUNE 2005 17

 REVIEW
target protein will produce the modified protein when supplemented
with the unnatural amino acid.
The unnatural amino acid mutagenesis method has been used to
introduce chemical reporter groups into proteins in both E. coli20,89–91
and yeast92,93 (Fig. 5b). For example, m-acetylphenylalanine was site-
specifically incorporated into LamB, an outer-membrane protein of
E. coli, and subsequently labeled with membrane-impermeant hydra-
zide dyes20. Similarly, azido and alkynyl amino acids related to tyro-
sine were installed in proteins within both E. coli and yeast90,93. After
cell lysis, the derivatized proteins were tagged by copper-catalyzed
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
[3+2] cycloaddition.
The above methods use the cell’s translational machinery to incorpo-
rate bioorthogonal functionality into proteins. Cells also possess a rich
machinery for post-translational modification that might be exploited
for similar purposes. This notion was recently explored by Ting and
coworkers using E. coli biotin ligase (BirA), an enzyme capable of attach-
ing a biotin prosthetic group to a 15-residue consensus sequence19.
BirA can recognize this sequence irrespective of its surrounding context
and can also tolerate subtle modification to the biotin structure. These
features were combined in a general tagging strategy wherein the con-
sensus sequence served as a gene-encoded tag for enzymatic ligation
of a keto-biotin analog (Fig. 5c). The ketone could then be modified
with fluorescent hydrazide probes. Further engineering of BirA might
enable enzymatic transfer of biotin analogs bearing other chemical
reporters. The method is technically straightforward and potentially
generalizable across a broad range of proteins and cell types. Other
strategies for enzymatic labeling of a target peptide sequence have
also been reported recently, and these might be considered alternative
avenues for the delivery of chemical reporters to proteins94,95.
The direct enzymatic conversion of an amino acid side chain to a
chemical reporter would be an appealing means for site-specific protein
labeling. An opportunity to achieve this was recently presented by the
discovery of the formylglycine-generating enzyme (FGE)96. Responsible
for converting sulfatases from an inactive to an active state, this enzyme
converts a critical cysteine residue to formylglycine (bearing an alde-
hyde at the C-α position) within a conserved 13-residue consensus
sequence. Like BirA, FGE will modify its target sequence irrespective
of the surrounding context. Thus, the FGE consensus sequence can be
imported into heterologous proteins and function as a general ‘alde-
hyde tag’ for subsequent chemical labeling with aminooxy or hydrazide
reagents (Fig. 5c and C.R.B., unpublished data).
Glycans and glycoconjugates. A powerful feature of chemical report-
ers is their applicability to labeling not just proteins but many classes of
biopolymers. Indeed, chemical reporters may turn out to be as broadly
applicable for labeling of glycans, lipids and other metabolites as GFP
has been for proteins. In the field of glycobiology, ketones and azides
have already proven to be useful markers for visualizing glycans within
their native environment. These biopolymers are known to mediate cell
surface recognition events97 and intracellular trafficking98 and, in recent
years, have also been implicated in transcriptional regulation99,100.
Glycans can participate in direct interactions with receptors, or they can
exert their biological activities in an indirect fashion by modulating the
functions of the proteins or lipids to which they are attached. The ability
to monitor glycans both independently of, and in conjunction with, the
scaffold to which they are attached could provide fundamental insights
into their roles in cell biology. At the cellular level, changes in glycosy-
lation are known to correlate with malignant transformation101 and the
development of a chronic inflammatory state102. Visualization of these
changes at the level of glycan structures would add a new dimension to
our understanding of the underlying pathology.
18
Chemical reporters can be embedded within glycans using endo-
genous biosynthetic pathways (a process we have previously termed
metabolic oligosaccharide engineering), then elaborated with small-
molecule probes for detection or isolation103. This two-step tagging
scheme has been used to study glycoconjugates containing the mono-
saccharides sialic acid (Sia), N-acetylgalactosamine (GalNAc) and
N-acetylglucosamine (GlcNAc).
The sialic acid biosynthetic pathway is permissive of unnatural N-acyl
substituents, and this site has been identified as suitable for the addition
of chemical reporters103,104. Mahal et al. reported that mammalian cells
metabolize the precursor sugar N-levulinoylmannosamine (ManLev),
an unnatural keto analog of N-acetylmannosamine (ManNAc), to the
corresponding keto sialic acid (SiaLev) on cell surface glycans22. The
unnatural residues can be tagged with a variety of probes, including
fluorophores and MRI contrast reagents105. Similarly, azides have been
incorporated into cell surface glycans by metabolism of the unnatu-
ral azido sugar N-azidoacetylmannosamine (ManNAz) (Fig. 6)32. The
resulting azido sialic acid (SiaNAz) residues can be labeled with various
probes through Staudinger ligation with phosphines32 or [3+2] cyclo-
addition with alkynes34. In addition, free sialic acid analogs themselves
can be used to deliver both ketones and azides to cell surface sialo-
glycoconjugates23. These intermediates enter the metabolic pathway
downstream of the corresponding mannosamines. They bypass the
most restrictive enzymes in the pathway and can therefore be adorned
with larger chemical reporters (such as aryl azides).
In addition to the sialoside biosynthetic pathway, both the GalNAc
and GlcNAc salvage pathways are tolerant of unnatural sugars bearing
bioorthogonal functionality. By this route, keto106 and azido64 GalNAc
analogs (such as N-azidoacetylgalactosamine or GalNAz (Fig. 6) can be
incorporated into mucin-type O-linked glycoproteins. Similarly, cells
incubated with N-azidoacetylglucosamine (GlcNAz, Fig. 6) will append
the unnatural residue to cytosolic and nuclear proteins at sites normally
occupied by O-GlcNAc65. The azide-labeled glycoproteins can be tagged
in a secondary reaction with fluorescent probes or affinity tags, enabling
them to be visualized and enriched from complex cell and tissue lysates.
In the future, such experiments might be used to inventory changes in
glycoprotein profiles in normal versus diseased cells.
Lipids and other biomolecules. Lipids are a class of biomolecules
that serve critical roles in cellular function and can also modulate the
activities of other biomolecules, such as proteins and glycans. Like gly-
cans, they cannot be readily labeled using genetic methods. Chemical
reporters incorporated into lipids by metabolism of their biosynthetic
precursors could provide the means to inventory and image these bio-
molecules within their native environments. The first example of this
was demonstrated in a proteomic analysis of farnesylated proteins25.
Protein farnesylation is a post-translational modification that can
relocalize cytosolic proteins to membranes, modulate protein confor-
mational changes and potentially mediate protein-protein interac-
tions107. The complete repertoire of farnesylated proteins is not known
and is difficult to predict by analysis of primary sequence alone. Zhao
and coworkers applied the bioorthogonal chemical reporter strategy
to label farnesylated proteins in cells for subsequent enrichment and
proteomic analysis25. They incubated cells with azido analogs of either
farnesol or farnesyl pyrophosphate and then tagged the modified pro-
teins by reaction of cell lysates with a biotin-derivatized phosphine.
Purification by avidin capture followed by mass spectrometric analysis
revealed several farnesylated proteins involved in a variety of processes,
including nucleosome assembly and peroxisome biogenesis. A similar
approach might be applied to profiling proteins modified with other
lipid groups such as geranylgeranyl, palmitoyl or myristoyl moieties.
VOLUME 1 NUMBER 1 JUNE 2005 NATURE CHEMICAL BIOLOGY
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
Figure 6 Azides can be incorporated into glycoconjugates using glycan biosynthetic pathways. Azido analogs of ManNAc (ManNAz) and sialic acid (SiaNAz)
are metabolized by cells and converted to cell surface azido sialosides. Similarly, an azido analog of GalNAc (GalNAz) can be metabolically introduced at the
core position of mucin-type O-linked glycoproteins. An azido analog of GlcNAc (GlcNAz) can be incorporated into cytosolic and nuclear glycoproteins.
In principle, any biomolecule can be studied using the bioorthogo-
nal chemical reporter strategy, as long as its biosynthetic pathway
is tolerant of modified precursors. Azide-modified nucleotides have
been incorporated into nucleic acids both in vitro and in living cells to
study protein-DNA and DNA-DNA interactions60,108. Thus, a simple
extrapolation indicates that chemical reporters could be incorporated
into nucleic acids within living cells and covalently labeled with chem-
ical probes. We anticipate that the bioorthogonal chemical reporter
strategy will also find utility in the profiling of other biomolecules
(such as cofactors) and post-translational modifications (such as
acetylation or methylation) in living systems. Indeed, syntheses of
azide-bearing flavonoids109 and S-adenosylmethionine derivatives
have recently been reported110,111.
Chemical reporters as readouts of enzyme function
In addition to their use in monitoring biomolecule expression and
localization, chemical reporters can provide a readout of enzyme func-
tion. In this case, the target protein is labeled with the chemical reporter
by virtue of its catalytic activity on a modified substrate, rather than
through the cell’s metabolic machinery. Termed activity-based protein
profiling by Cravatt and coworkers, this approach has been used to
identify active glutathione S-transferases26, glycosidases28 and protea-
some molecules27. In each case, a mechanism-based covalent inhibitor
of the target protein class was designed to incorporate the azide group.
Catalytically active proteins were covalently labeled with the inhibi-
tor and then selectively tagged with phosphine- or alkyne-modified
probes, permitting analysis by western blotting or enrichment for mass
spectrometric analysis. This approach to activity-based labeling can be
applied, in principle, to any enzyme class for which a selective covalent
modifier is available. In a broader profiling experiment, overall levels
NATURE CHEMICAL BIOLOGY VOLUME 1 NUMBER 1 JUNE 2005
REVIEW
of enzymes with active-site nucleophiles were compared in various
breast cancer cell lines29. In this case, an electrophilic substrate bearing
an alkyne reporter was found to give cleaner labeling than the corre-
sponding azido analog.
Bioorthogonal chemical reporters in living organisms
One of the most dramatic applications of GFP-protein fusions has been
noninvasive imaging of protein expression and localization in living
organisms ranging from Caenorhabditis elegans to mice4,112. Chemical
reporters might provide similar opportunities for other classes of bio-
molecules. Already, both proteins and glycans have been labeled with
azides in laboratory mice, using covalent enzyme inhibitors26 and azido
sugars30, respectively. For noninvasive imaging applications, the sec-
ondary tagging reaction with a phosphine or alkyne probe must also
be accomplished in the living organism.
The demands on bioorthogonal reactions in this context are far
more stringent than those for cellular systems. Aside from having
extraordinary chemical selectivity, the reagents must not be prone to
rapid metabolic breakdown or excretion, and they must not accumu-
late in cells or organs nonspecifically on the timescale of the reaction.
Very few covalent chemistries have been relocated from the round-
bottom flask to a living organism. We recently investigated whether
cells labeled with azido sugars in laboratory mice were capable of
further chemical modification by Staudinger ligation30. Mice were
injected with either ManNAz or GalNAz for several days and then
administered a phosphine probe. After several hours, the anticipated
product of the Staudinger ligation was observed on splenocyte cell
surfaces and serum glycoproteins (C.R.B., unpublished data, ref. 30).
In the future, this chemical reporter–bioorthogonal reaction system
might enable noninvasive imaging of glycan expression.
19
 REVIEW

Summary and future outlook 6.
The bioorthogonal chemical reporter strategy offers a means to visual-
7.
ize multiple classes of biomolecules in living systems. Substrates linked
to chemical reporters can be metabolized by cells and incorporated 8.
into proteins, glycans, lipids and other cellular species. After covalent 9.
reaction with complementary probes, these classes of biomolecules can
be visualized in living cells and, in some cases, living organisms. Both
10.
discrete biopolymers (such as tetracysteine-fused proteins) and entire
subsets of biomolecules (such as metabolically labeled glycans or lipids) 11.
can be tracked in their native habitats using this technology. 12.
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
Presley, J.F. et al. ER-to-Golgi transport visualized in living cells. Nature 389, 81–85
(1997).
Dundr, M. et al. A kinetic framework for a mammalian RNA polymerase in vivo.
Science 298, 1623–1626 (2002).
Miyawaki, A., Sawano, A. & Kogure, T. Lighting up cells: labelling proteins with
fluorophores. Nat. Cell Biol. 5 (Suppl.), S1–S7 (2003).
Shaner, N.C. et al. Improved monomeric red, orange and yellow fluorescent pro-
teins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 22,
1567–1572 (2004).
Verkhusha, V.V. & Lukyanov, K.A. The molecular properties and applications of Anthozoa
fluorescent proteins and chromoproteins. Nat. Biotechnol. 22, 289–296 (2004).
Alberts, B. Molecular Biology of the Cell (Garland Science, New York, 2002)
Schweppe, R.E., Haydon, C.E., Lewis, T.S., Resing, K.A. & Ahn, N.G. The charac-

terization of protein post-translational modifications by mass spectrometry. Acc.

Several challenges remain with respect to both metabolic labeling
and chemical tagging in biological systems. In many cases, competition
with endogenous substrates, such as natural amino acids, lipids and
sugars, cannot be avoided. In such cases, one should expect metabolic
substitution with the reporter-modified building block to be incom-
plete. The fraction of biomolecules labeled with the chemical reporter
may be an important parameter when interpreting results and therefore
must be quantified in some circumstances. Another consideration is the
physiological consequence associated with the addition of a chemical
reporter to a biomolecule, particularly in living organisms. Even subtle
perturbations to the structure of a protein, glycan or lipid may affect
its biological activity, localization or stability. A third issue to address is
the problem of slow reaction kinetics, which can undermine the use of
bioorthogonal reactions for biomolecule tagging. The chemical labeling
step involves a bimolecular reaction with a second-order rate constant
that is typically far below that of a noncovalent binding event (such as
an antibody-antigen interaction). Rapid reactions are essential for the
observation of biological events that occur on a very short time scale or
among biomolecules of low abundance. This problem might be solved
by engineering a fast, reversible association of the reaction partners that
precedes an irreversible covalent labeling step.
Even at this early point in the development of the technique, it is
clear that chemical reporters and bioorthogonal reactions have a rich
future in the field of chemical biology. Some interesting future direc-
tions include imposing temporal and spatial control over metabolic
labeling with the reporter functional group. This might be accom-
plished using caged substrates that are released by light- or tissue-
specific enzymes. A future challenge for synthetic chemists will be to
craft novel bioorthogonal transformations for use in living organisms.
Just as combinations of fluorescent proteins (such as CFP and YFP)
have proven useful in studying multicomponent processes, an arsenal
of bioorthogonal reactions could find use in monitoring collections of
species that function together in living systems.
ACKNOWLEDGMENTS
J.A.P. is supported by a Howard Hughes Medical Institute predoctoral fellowship.
We thank N. Agard, J. Baskin, I. Carrico, D. Dube, S. Laughlin and C. McVaugh for
critical reading of the manuscript.
COMPETING INTERESTS STATEMENT
The authors delare that they have no competing financial interests.
Published online at http://www.nature.com/naturechemicalbiology/
1. Tsien, R.Y. The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544
(1998).
2. Lippincott-Schwartz, J. & Patterson, G.H. Development and use of fluorescent pro-
tein markers in living cells. Science 300, 87–91 (2003).
3. Zhang, J., Campbell, R.E., Ting, A.Y. & Tsien, R.Y. Creating new fluorescent probes
for cell biology. Nat. Rev. Mol. Cell Biol. 3, 906–918 (2002).
4. Hadjantonakis, A.K., Dickinson, M.E., Fraser, S.E. & Papaioannou, V.E. Technicolour
transgenics: imaging tools for functional genomics in the mouse. Nat. Rev. Genet.
4, 613–625 (2003).
5. Weissleder, R. & Ntziachristos, V. Shedding light onto live molecular targets. Nat.
Med. 9, 123–128 (2003).
Chem. Res. 36, 453–461 (2003).
13. Massoud, T.F. & Gambhir, S.S. Molecular imaging in living subjects: seeing funda-
mental biological processes in a new light. Genes Dev. 17, 545–580 (2003).
14. von Mehren, M., Adams, G.P. & Weiner, L.M. Monoclonal antibody therapy for cancer.
Annu. Rev. Med. 54, 343–369 (2003).
15. Hudson, P.J. & Souriau, C. Engineered antibodies. Nat. Med. 9, 129–134
(2003).
16. Hermanson, G.T. Bioconjugate Techniques (Academic Press, San Diego, 1996).
17. Link, A.J., Vink, M.K. & Tirrell, D.A. Presentation and detection of azide functionality
in bacterial cell surface proteins. J. Am. Chem. Soc. 126, 10598–10602 (2004).
18. Griffin, B.A., Adams, S.R. & Tsien, R.Y. Specific covalent labeling of recombinant
protein molecules inside live cells. Science 281, 269–272 (1998).
19. Chen, I., Howarth, M., Lin, W. & Ting, A.Y. Site-specific labeling of cell surface pro-
teins with biophysical probes using biotin ligase. Nat. Methods 2, 99–104 (2005).
20. Zhang, Z. et al. A new strategy for the site-specific modification of proteins in vivo.
Biochemistry 42, 6735–6746 (2003).
21. Saxon, E. et al. Investigating cellular metabolism of synthetic azidosugars with the
Staudinger ligation. J. Am. Chem. Soc. 124, 14893–14902 (2002).
22. Mahal, L.K., Yarema, K.J. & Bertozzi, C.R. Engineering chemical reactivity on cell
surfaces through oligosaccharide biosynthesis. Science 276, 1125–1128 (1997).
23. Luchansky, S.J., Goon, S. & Bertozzi, C.R. Expanding the diversity of unnatural
cell-surface sialic acids. ChemBioChem 5, 371–374 (2004).
24. Luchansky, S.J. et al. Constructing azide-labeled cell surfaces using polysaccharide
biosynthetic pathways. Methods Enzymol. 362, 249–272 (2003).
25. Kho, Y. et al. A tagging-via-substrate technology for detection and proteomics of
farnesylated proteins. Proc. Natl. Acad. Sci. USA 101, 12479–12484 (2004).
26. Speers, A.E., Adam, G.C. & Cravatt, B.F. Activity-based protein profiling in vivo
using a copper(I)-catalyzed azide-alkyne [3 + 2] cycloaddition. J. Am. Chem. Soc.
125, 4686–4687 (2003).
27. Ovaa, H. et al. Chemistry in living cells: detection of active proteasomes by a two-
step labeling strategy. Angew. Chem. Int. Edn. Engl. 42, 3626–3629 (2003).
28. Vocadlo, D.J. & Bertozzi, C.R. A strategy for functional proteomic analysis of glycosidase
activity from cell lysates. Angew. Chem. Int. Edn. Engl. 43, 5338–5342 (2004).
29. Speers, A.E. & Cravatt, B.F. Profiling enzyme activities in vivo using click chemistry
methods. Chem. Biol. 11, 535–546 (2004).
30. Prescher, J.A., Dube, D.H. & Bertozzi, C.R. Chemical remodelling of cell surfaces
in living animals. Nature 430, 873–877 (2004).
31. Adams, S.R. et al. New biarsenical ligands and tetracysteine motifs for protein
labeling in vitro and in vivo: synthesis and biological applications. J. Am. Chem.
Soc. 124, 6063–6076 (2002).
32. Saxon, E. & Bertozzi, C.R. Cell surface engineering by a modified Staudinger reac-
tion. Science 287, 2007–2010 (2000).
33. Kolb, H.C. & Sharpless, K.B. The growing impact of click chemistry on drug discov-
ery. Drug Discov. Today 8, 1128–1137 (2003).
34. Agard, N.J., Prescher, J.A. & Bertozzi, C.R. A strain-promoted [3 + 2] azide-alkyne
cycloaddition for covalent modification of biomolecules in living systems. J. Am.
Chem. Soc. 126, 15046–15047 (2004).
35. Andresen, M., Schmitz-Salue, R. & Jakobs, S. Short tetracysteine tags to beta-
tubulin demonstrate the significance of small labels for live cell imaging. Mol. Biol.
Cell 15, 5616–5622 (2004).
36. Panchal, R.G. et al. In vivo oligomerization and raft localization of Ebola virus protein
VP40 during vesicular budding. Proc. Natl. Acad. Sci. USA 100, 15936–15941
(2003).
37. Gaietta, G. et al. Multicolor and electron microscopic imaging of connexin traffick-
ing. Science 296, 503–507 (2002).
38. Ju, W. et al. Activity-dependent regulation of dendritic synthesis and trafficking of
AMPA receptors. Nat. Neurosci. 7, 244–253 (2004).
39. Tour, O., Meijer, R.M., Zacharias, D.A., Adams, S.R. & Tsien, R.Y. Genetically targeted
chromophore-assisted light inactivation. Nat. Biotechnol. 21, 1505–1508 (2003).
40. Chen, I. & Ting, A.Y. Site-specific labeling of proteins with small molecules in live
cells. Curr. Opin. Biotechnol. 16, 35–40 (2005).
41. Hahn, M.E. & Muir, T.W. Manipulating proteins with chemistry: a cross-section of
chemical biology. Trends Biochem. Sci. 30, 26–34 (2005).
42. Johnsson, N. & Johnsson, K. A fusion of disciplines: chemical approaches to exploit
fusion proteins for functional genomics. ChemBioChem 4, 803–810 (2003).
43. van Swieten, P.F., Leeuwenburgh, M.A., Kessler, B.M. & Overkleeft, H.S.
Bioorthogonal organic chemistry in living cells: novel strategies for labeling bio-
molecules. Org. Biomol. Chem. 3, 20–27 (2005).
44. Keppler, A., Pick, H., Arrivoli, C., Vogel, H. & Johnsson, K. Labeling of fusion

20 VOLUME 1 NUMBER 1 JUNE 2005 NATURE CHEMICAL BIOLOGY


proteins with synthetic fluorophores in live cells. Proc. Natl. Acad. Sci. USA 101,
9955–9959 (2004).
45. Marks, K.M., Braun, P.D. & Nolan, G.P. A general approach for chemical labeling
and rapid, spatially controlled protein inactivation. Proc. Natl. Acad. Sci. USA 101,
9982–9987 (2004).
46. Miller, L.W., Sable, J., Goelet, P., Sheetz, M.P. & Cornish, V.W. Methotrexate conju-
gates: a molecular in vivo protein tag. Angew. Chem. Int. Edn. Engl. 43, 1672–1675
(2004).
47. Miller, L.W., Cai, Y., Sheetz, M.P. & Cornish, V.W. In vivo protein labeling with trim-
ethoprim conjugates: a flexible chemical tag. Nat. Methods 2, 255–257 (2005).
48. Guignet, E.G., Hovius, R. & Vogel, H. Reversible site-selective labeling of membrane
proteins in live cells. Nat. Biotechnol. 22, 440–444 (2004).
49. Marks, K.M., Rosinov, M. & Nolan, G.P. In vivo targeting of organic calcium sensors
© 2005 Nature Publishing Group http://www.nature.com/naturechemicalbiology
via genetically selected peptides. Chem. Biol. 11, 347–356 (2004).
50. Franz, K.J., Nitz, M. & Imperiali, B. Lanthanide-binding tags as versatile protein
coexpression probes. ChemBioChem 4, 265–271 (2003).
51. Giriat, I. & Muir, T.W. Protein semi-synthesis in living cells. J. Am. Chem. Soc. 125,
7180–7181 (2003).
52. Jencks, W.P. Studies on the mechanism of oxime and semicarbazone formation.
J. Am. Chem. Soc. 81, 475–481 (1959).
53. Rideout, D. Self-assembling cytotoxins. Science 233, 561–563 (1986).
54. Rideout, D. Self-assembling drugs: a new approach to biochemical modulation in
cancer chemotherapy. Cancer Invest. 12, 189–202 (1994).
55. Rideout, D., Calogeropoulou, T., Jaworski, J. & McCarthy, M. Synergism through
direct covalent bonding between agents: a strategy for rational design of chemo-
therapeutic combinations. Biopolymers 29, 247–262 (1990).
56. Rotenberg, S.A., Calogeropoulou, T., Jaworski, J., Weinstein, I.B. & Rideout, D. A
self-assembling protein kinase C inhibitor. Proc. Natl. Acad. Sci. USA 88, 2490–
2494 (1991).
57. Yarema, K.J., Mahal, L.K., Bruehl, R.E., Rodriguez, E.C. & Bertozzi, C.R. Metabolic
delivery of ketone groups to sialic acid residues. Application to cell surface glycoform
engineering. J. Biol. Chem. 273, 31168–31179 (1998).
58. Lee, J.H. et al. Engineering novel cell surface receptors for virus-mediated gene
transfer. J. Biol. Chem. 274, 21878–21884 (1999).
59. Sadamoto, R. et al. Control of bacteria adhesion by cell-wall engineering. J. Am.
Chem. Soc. 126, 3755–3761 (2004).
60. Griffin, R.J. The medicinal chemistry of the azido group. Prog. Med. Chem. 31,
121–232 (1994).
61. Saegusa, T., Ito, Y. & Shimizu, T. Synthetic reactions by complex catalysts. XVII. Copper-
catalyzed reaction of azide with thiol. J. Org. Chem. 35, 2979–2981 (1970).
62. Staudinger, H. & Meyer, J. Uber neue organische phosphoverbindungen III.
Phosphinmethlenderivate und phosphinimine. Helv. Chim. Acta 2, 635–646 (1919).
63. Gololobov, Y.G. & Kasukhin, L.F. Recent advances in the Staudinger reaction.
Tetrahedron 48, 1353–1406 (1992).
64. Hang, H.C., Yu, C., Kato, D.L. & Bertozzi, C.R. A metabolic labeling approach toward
proteomic analysis of mucin-type O-linked glycosylation. Proc. Natl. Acad. Sci. USA
100, 14846–14851 (2003).
65. Vocadlo, D.J., Hang, H.C., Kim, E.J., Hanover, J.A. & Bertozzi, C.R. A chemical
approach for identifying O-GlcNAc-modified proteins in cells. Proc. Natl. Acad. Sci.
USA 100, 9116–9121 (2003).
66. Luchansky, S.J., Argade, S., Hayes, B.K. & Bertozzi, C.R. Metabolic functionalization
of recombinant glycoproteins. Biochemistry 43, 12358–12366 (2004).
67. Nilsson, B.L., Kiessling, L.L. & Raines, R.T. Staudinger ligation: a peptide from a
thioester and azide. Org. Lett. 2, 1939–1941 (2000).
68. Saxon, E., Armstrong, J.I. & Bertozzi, C.R.A. “Traceless” Staudinger ligation for the
chemoselective synthesis of amide bonds. Org. Lett. 2, 2141–2143 (2000).
69. Kohn, M. et al. Staudinger ligation: a new immobilization strategy for the preparation
of small-molecule arrays. Angew. Chem. Int. Edn. Engl. 42, 5830–5834 (2003).
70. Soellner, M.B., Dickson, K.A., Nilsson, B.L. & Raines, R.T. Site-specific protein immo-
bilization by Staudinger ligation. J. Am. Chem. Soc. 125, 11790–11791 (2003).
71. Kohn, M. & Breinbauer, R. The Staudinger ligation—a gift to chemical biology.
Angew. Chem. Int. Edn. Engl. 43, 3106–3116 (2004).
72. Huisgen, R. 1,3-Dipolar Cycloaddition Chemistry (ed. Padwa, A.) (Wiley, New York,
1984).
73. Huisgen, R. 1,3-Dipolar cycloadditions. Angew. Chem. Int. Edn. Engl. 2, 565–598
(1963).
74. Rostovtsev, V.V., Green, L.G., Fokin, V.V. & Sharpless, K.B. A stepwise Huisgen
cycloaddition process: copper(I)-catalyzed regioselective “ligation” of azides and
terminal alkynes. Angew. Chem. Int. Edn. Engl. 41, 2596–2599 (2002).
75. Tornoe, C.W., Christensen, C. & Meldal, M. Peptidotriazoles on solid phase: [1,2,3]-
triazoles by regiospecific copper(I)-catalyzed 1,3-dipolar cycloadditions of terminal
alkynes to azides. J. Org. Chem. 67, 3057–3064 (2002).
76. Wang, Q. et al. Bioconjugation by copper(I)-catalyzed azide-alkyne [3 + 2] cycload-
dition. J. Am. Chem. Soc. 125, 3192–3193 (2003).
77. Seo, T.S. et al. Photocleavable fluorescent nucleotides for DNA sequencing on a
chip constructed by site-specific coupling chemistry. Proc. Natl. Acad. Sci. USA
101, 5488–5493 (2004).
78. Speers, A.E. & Cravatt, B.F. Chemical strategies for activity-based proteomics.
ChemBioChem 5, 41–47 (2004).
79. Link, A.J. & Tirrell, D.A. Cell surface labeling of Escherichia coli via copper(I)-cata-
lyzed [3+2] cycloaddition. J. Am. Chem. Soc. 125, 11164–11165 (2003).
80. Turner, R., Jarrett, A.D., Goebel, P. & Mallon, B.J. Heats of hydrogenation. IX.
Cyclic acetylenes and some miscellaneous olefins. J. Am. Chem. Soc. 95, 790–792
(1972).
NATURE CHEMICAL BIOLOGY VOLUME 1 NUMBER 1 JUNE 2005
REVIEW
81. Wittig, G.A.K.A. Zur Existenz niedergliedriger Cycloalkine, I. Chem. Ber. 94, 3260–
3275 (1961).
82. Lin, F.L., Hoyt, H.M., van Halbeek, H., Bergman, R.G. & Bertozzi, C.R. Mechanistic
investigation of the Staudinger ligation. J. Am. Chem. Soc. 127, 2686–2695
(2005).
83. Link, A.J., Mock, M.L. & Tirrell, D.A. Non-canonical amino acids in protein engineer-
ing. Curr. Opin. Biotechnol. 14, 603–609 (2003).
84. Wang, L. & Schultz, P.G. Expanding the genetic code. Angew. Chem. Int. Edn. Engl.
44, 34–66 (2004).
85. Datta, D., Wang, P., Carrico, I.S., Mayo, S.L. & Tirrell, D.A. A designed phenylala-
nyl-tRNA synthetase variant allows efficient in vivo incorporation of aryl ketone
functionality into proteins. J. Am. Chem. Soc. 124, 5652–5653 (2002).
86. Kirshenbaum, K., Carrico, I.S. & Tirrell, D.A. Biosynthesis of proteins incorporating
a versatile set of phenylalanine analogues. ChemBioChem 3, 235–237 (2002).
87. van Hest, J.C.M., Kiick, K.L. & Tirrell, D.A. Efficient incorporation of unsaturated methi-
onine analogs into proteins in vivo. J. Am. Chem. Soc. 122, 1282–1288 (2000).
88. Kiick, K.L., Saxon, E., Tirrell, D.A. & Bertozzi, C.R. Incorporation of azides into
recombinant proteins for chemoselective modification by the Staudinger ligation.
Proc. Natl. Acad. Sci. USA 99, 19–24 (2002).
89. Chin, J.W. et al. Addition of p-azido-L-phenylalanine to the genetic code of
Escherichia coli. J. Am. Chem. Soc. 124, 9026–9027 (2002).
90. Deiters, A. & Schultz, P.G. In vivo incorporation of an alkyne into proteins in
Escherichia coli. Bioorg. Med. Chem. Lett. 15, 1521–1524 (2005).
91. Wang, L., Zhang, Z., Brock, A. & Schultz, P.G. Addition of the keto functional group
to the genetic code of Escherichia coli. Proc. Natl. Acad. Sci. USA 100, 56–61
(2003).
92. Chin, J.W. et al. An expanded eukaryotic genetic code. Science 301, 964–967
(2003).
93. Deiters, A. et al. Adding amino acids with novel reactivity to the genetic code of
Saccharomyces cerevisiae. J. Am. Chem. Soc. 125, 11782–11783 (2003).
94. George, N., Pick, H., Vogel, H., Johnsson, N. & Johnsson, K. Specific labeling of
cell surface proteins with chemically diverse compounds. J. Am. Chem. Soc. 126,
8896–8897 (2004).
95. Yin, J., Liu, F., Li, X. & Walsh, C.T. Labeling proteins with small molecules by site-spe-
cific posttranslational modification. J. Am. Chem. Soc. 126, 7754–7755 (2004).
96. Dierks, T. et al. Multiple sulfatase deficiency is caused by mutations in the gene encod-
ing the human C(α)-formylglycine generating enzyme. Cell 113, 435–444 (2003).
97. Rudd, P.M., Elliott, T., Cresswell, P., Wilson, I.A. & Dwek, R.A. Glycosylation and
the immune system. Science 291, 2370–2376 (2001).
98. Gouyer, V. et al. Inhibition of the glycosylation and alteration in the intracellular
trafficking of mucins and other glycoproteins by GalNAcα-O-Bn in mucosal cell lines:
an effect mediated through the intracellular synthesis of complex GalNAcα-O-Bn
oligosaccharides. Front. Biosci. 6, D1235–D1244 (2001).
99. Wells, L., Vosseller, K. & Hart, G.W. Glycosylation of nucleocytoplasmic proteins:
signal transduction and O-GlcNAc. Science 291, 2376–2378 (2001).
100. Slawson, C. & Hart, G.W. Dynamic interplay between O-GlcNAc and O-phosphate: the
sweet side of protein regulation. Curr. Opin. Struct. Biol. 13, 631–636 (2003).
101. Orntoft, T.F. & Vestergaard, E.M. Clinical aspects of altered glycosylation of glyco-
proteins in cancer. Electrophoresis 20, 362–371 (1999).
102. Lowe, J.B. Glycan-dependent leukocyte adhesion and recruitment in inflammation.
Curr. Opin. Cell Biol. 15, 531–538 (2003).
103. Dube, D.H. & Bertozzi, C.R. Metabolic oligosaccharide engineering as a tool for
glycobiology. Curr. Opin. Chem. Biol. 7, 616–625 (2003).
104. Keppler, O.T., Horstkorte, R., Pawlita, M., Schmidt, C. & Reutter, W. Biochemical
engineering of the N-acyl side chain of sialic acid: biological implications.
Glycobiology 11, 11R–18R (2001).
105. Lemieux, G.A. & Bertozzi, C.R. Exploiting differences in sialoside expression for
selective targeting of MRI contrast reagents. J. Am. Chem. Soc. 121, 4278–4279
(1999).
106. Hang, H.C. & Bertozzi, C.R. Ketone isosteres of 2-N-acetamidosugars as sub-
strates for metabolic cell surface engineering. J. Am. Chem. Soc. 123, 1242–1243
(2001).
107. Zhang, F.L. & Casey, P.J. Protein prenylation: molecular mechanisms and functional
consequences. Annu. Rev. Biochem. 65, 241–269 (1996).
108. Sylvers, L.A. & Wower, J. Nucleic acid-incorporated azidonucleotides: probes for
studying the interaction of RNA or DNA with proteins and other nucleic acids.
Bioconjug. Chem. 4, 411–418 (1993).
109. Malolanarasimhan, K. et al. Synthesis and biological study of a flavone acetic acid
analogue containing an azido reporting group designed as a multifunctional binding
site probe. Bioorg. Med. Chem. 13, 2717–2722 (2005).
110. Comstock, L.R. & Rajski, S.R. Efficient synthesis of azide-bearing cofactor mimics.
J. Org. Chem. 69, 1425–1428 (2004).
111. Comstock, L.R. & Rajski, S.R. Conversion of DNA methyltransferases into azido-
nucleosidyl transferases via synthetic cofactors. Nucleic Acids Res. 33, 1644–1652
(2005).
112. Poteryaev, D., Squirrell, J.M., Campbell, J.M., White, J.G. & Spang, A. Involvement
of the actin cytoskeleton and homotypic membrane fusion in ER dynamics in
C. elegans. Mol. Biol. Cell (2005).
113. Apweiler, R., Hermjakob, H. & Sharon, N. On the frequency of protein glycosylation,
as deduced from analysis of the SWISS-PROT database. Biochim. Biophys. Acta
1473, 4–8 (1999).
114. Restituyo, J.A., Comstock, L.R., Petersen, S.G., Stringfellow, T. & Rajski, S.R.
Conversion of aryl azides to O-alkyl imidates via modified Staudinger ligation.
Org. Lett. 5, 4357–4360 (2003).
21