A fluorogenic dye activated by the Staudinger ligation

Article abstract of DOI:10.1021/ja029013y

Specific labeling of biomolecules with biochemical and biophysical probes is a central element of proteomics research. Here we describe a coumarin-phosphine dye that undergoes activation of coumarin fluorescence upon Staudinger ligation with azides. Since azides can be metabolically incorporated into cellular proteins and oligosaccharides, this dye may be a useful tool for profiling proteins and their posttranslational modifications. Copyright

Full text of DOI:10.1021/ja029013y

Published on Web 03/28/2003
A Fluorogenic Dye Activated by the Staudinger Ligation
George A. Lemieux, Christopher L. de Graffenried, and Carolyn R. Bertozzi*
Center for New Directions in Organic Synthesis, Departments of Chemistry and Molecular and Cell Biology, and
Howard Hughes Medical Institute, UniVersity of California, Berkeley, California 94720
Received October 18, 2002; E-mail: bertozzi@cchem.berkeley.edu
In the emerging field of proteomics, the demands of parallel
purification, quantitation, and identification of proteins from
complex cellular extracts has necessitated the development of
methods for their selective labeling with detectable probes.1-3
Posttranslational modifications of proteins such as glycosylation
and lipidation introduce significant additional complexity into the
proteome.4 While the protein itself may be labeled with gene-
encoded tags such as fluorescent proteins5 and FLAsH tetracysteine
motifs,6 there are few general tools available for labeling post-
translational modifications.
Previous work in our lab has established that unnatural sugar
substrates bearing bio-orthogonal functional groups can be meta-
bolically incorporated into cellular glycoproteins.7-10 Selective
covalent reaction of the functional groups with biochemical or
biophysical probes permits the detection and isolation of proteins
Figure 1. (A) The Staudinger ligation. (B) A coumarin-phosphine
fluorogenic dye (1) activated by the Staudinger ligation.
bearing the corresponding sugar as a posttranslational modification.
The azide is a sterically undemanding, bio-orthogonal electrophile
that can be introduced into myriad metabolic precursors.7 The
Staudinger ligation between azides and engineered phosphine
reagents (Figure 1A) has proven particularly useful for bio-
orthogonal labeling of glycans on mammalian glycoproteins,9 and
unnatural amino acids within proteins expressed in Escherichia
coli.10 The primary labeling reagents used in these experiments,
biotin9 and FLAG peptide-phosphines,10 were designed for detec-
tion using fluorescent secondary reagents (FITC-avidin and FITC-
anti-FLAG antibody, respectively). The two-step procedure per-
mitted removal of excess unbound primary label before the detection
step, and as a result, minimized the background signal that can
otherwise frustrate covalent detection experiments.
For some applications, however, it is either inconvenient or
simply impossible to use such two-step labeling procedures. For
example, excess primary labeling reagents cannot be easily washed
away from intracellular compartments or from tissues of living
organisms. If the ligation product could be readily distinguished
from the excess primary detection agent, the unbound agent would
not contribute to the background signal in an assay. A chemose-
lective reaction with this feature potentially obviates the need for
a secondary reagent since detection of the product could be achieved
in a single covalent labeling step.
Figure 2. (A) Synthesis of compound 1. (B) Detection of the fluorescence
of ligation product 5 in a microtiter plate excited at 365 nm.
rendering the molecule nonfluorescent. The oxidation of the
phosphine that occurs during the reaction eliminates quenching and
activates fluorescence.
Here we report a fluorogenic coumarin-phosphine dye (1, Figure
1B) that is activated by the Staudinger ligation with azides.
Compound 1 comprises a triaryl phosphine in which one phosphorus
ligand is a 2-substituted methyl benzoate that traps the intermediate
iminophosphorane to generate, after hydrolysis, an amide-linked
product.9 A second ligand on phosphorus is a 7-amino coumarin
attached via the 3-position. Substituents at this position of 7-amino
coumarin dyes are known to strongly influence their fluorescence
properties.11 Before the Staudinger ligation, the lone pair of electrons
on phosphorus quenches the excited state of the fluorophore,
The synthesis of compound 1 is described in detail in the
Supporting Information and is outlined in Figure 2A. Briefly, methyl
2-iodobenzoate (2) was converted to the Grignard reagent and then
condensed with one equivalent of dichlorophenylphosphine. Sub-
sequently, coumarin-zinc reagent 3 was added to generate com-
pound 1.
To test its reactivity and fluorescence properties, we reacted
compound 1 with model azide 4 in an acetonitrile/water12 mixture,
affording the ligated phosphine oxide 5 as the exclusive product
(Figure 2A). The increase in fluorescence during the reaction of 1
9
4708
J. AM. CHEM. SOC. 2003, 125, 4708-4709
10.1021/ja029013y CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Photophysical Parameters of Coumarin Analogues
ꢀ (M-1 cm-1 a
)
λ
abs (nm)a
λ
ex (nm)b
λ
em (nm)b
quantum yield
1
5
33,000 ( 2000
47,000 ( 2000
421
430
432
443
507
495
0.011 ( 0.001
0.65 ( 0.04
Figure 3. Specific labeling of azido-mDHFR with phosphine 1. Purified
azido-mDHFR (lanes 1 and 3) and native mDHRF (lanes 2 and 4) were
incubated overnight with 200 µM 1 in 40% EtOH/PBS, pH 7.1, and the
crude reaction was analyzed by 12% SDS-PAGE. Lanes 1 and 2 show the
silver-stained gel, indicating total protein content. Lanes 3 and 4 show the
same gel analyzed by fluorescence imaging (excitation at 457 nm, emission
collected with a 520 nm band-pass filter).
a Measured in ethanol. b Measured in phosphate-buffered saline (PBS).
and 4 was immediately evident and did not occur in control
reactions lacking 4 (Figure 2B).
The photophysical parameters of phosphine 1 and phosphine
oxide 5 are quite distinct and are summarized in Table 1. The
maximum absorbance of the ligated product 5 was slightly red-
shifted compared to phosphine 1. In addition, the molar absorptivity
of 5 was greater. The excitation spectra measured in phosphate-
buffered saline were observed to have maxima at 432 and 443 nm
for compounds 1 and 5, respectively. The most profound difference
between the two compounds was observed in their quantum yields
of fluorescence. Compound 5 showed intense emission with a
maximum at 495 nm and a quantum yield relative to quinine sulfate
of 0.65 ( 0.04. In contrast, phosphine 1 exhibited a very weak
fluorescence with a quantum yield of 0.011 ( 0.001 and an emission
maximum at 507 nm. Thus, reaction of phosphine 1 with azides
produces a product easily distinguishable from the reactants by its
intense fluorescence.
To demonstrate the utility of compound 1 for biomolecule
labeling, we reacted the phosphine dye with recombinant murine
dihydrofolate reductase bearing azidohomoalanine residues in place
of native methionine residues.10 This azido protein (azido-mDHFR)
was generated by metabolic incorporation of azidohomoalanine
during overexpression in a methionine auxotrophoic E. coli strain,
as previously reported.10 Azido-mDHFR and native mDHFR were
incubated with 200 µM phosphine 1, and the crude reactions were
loaded directly onto a gel with no separation of unreacted dye.
Analysis of the gel by silver stain (lanes 1 and 2) and fluorescence
imaging (lanes 3 and 4) showed specific labeling of the azido
protein by compound 1 and no detectable labeling of the native
protein lacking azides (Figure 3). Unlike previous experiments using
FLAG and biotin conjugates, the labeled protein could be directly
observed without need for Western blotting, washing, or secondary
labeling steps.
Figure 4. Kinetics of the reaction between compounds 1 and 4 under
pseudo-first-order conditions: (b) 5 mM 4, 0.011 mM 1; (9) 5 mM 4,
0.0022 mM 1; (×) 0.011 mM 1; (O) 0.0022 mM 1. Fluorescence intensities
were determined at λex ) 443 nm and λem ) 493 nm.
Acknowledgment. The Center for New Directions in Organic
Synthesis is supported by Bristol-Myers Squibb as Sponsoring
Member and Novartis as Supporting Member. We thank Eliana
Saxon, Sarah Luchansky, David Madden, and Roger Tsien for
technical advice, and David Tirrell and Kristi Kiick for azido-
mDHFR. G.A.L. was supported by NIH Biotechnology Training
Grant GM08352. This work was supported by NIH Grant GM58867.
Supporting Information Available: Synthetic procedures and
analytical data for compound 1 and its precursors, experimental details
for fluorescence quantum yield and kinetic measurements, excitation
and emission spectra for compounds 1 and 5, and procedures for protein
labeling (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
We determined the rate of reaction between phosphine 1 and
azide 4, exploiting fluorescence as a measure of product formation,
to optimize conditions for biomolecule labeling (Figure 4). In the
absence of azide 4 there was no significant change in the
fluorescence intensity with time, an indication that background
oxidation (presumably with dissolved molecular oxygen) occurs
at a relatively low rate. The rate data obtained in these experiments
were used to calculate an apparent second-order rate constant of
References
(1) Griffin, T. J.; Goodlett, D. R.; Aebersold, R. Curr. Opin. Biotechnol. 2001,
12, 607-612.
(2) Cravatt, B. F.; Sorensen, E. J. Curr. Opin. Chem. Biol. 2000, 4, 663-668.
(3) Moseley, M. A. Trends Biotechnol. 2001, 19, S10-S16.
(4) Mann, M.; Hendrickson, R. C.; Pandey, A. Annu. ReV. Biochem. 2001,
70, 437-473.
(5) Van Roessel. P.; Brand, A. H. Nat. Cell Biol. 2002, 4, E15-E20.
(6) Griffin, B. A.; Adams, S. R.; Tsien, R. Y. Science 1998, 281, 269-272.
(7) Saxon, E.; Bertozzi, C. R. Annu. ReV. Cell DeV. Biol. 2001, 17, 1-23.
(8) Yarema, K. J.; Goon, S.; Bertozzi, C. R. Nat. Biotechnol. 2001, 19, 553-
558.
(9) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007-2010.
(10) Kiick, K. L.; Saxon, E.; Tirrell, D. A.; Bertozzi, C. R. Proc. Natl. Acad.
Sci. U.S.A. 2002, 99, 19-24.
(11) Schiedel, M.-S.; Briehn, C. A.; Ba¨uerle, P. Angew. Chem., Int. Ed. 2001,
40, 4677-4680.
(12) The reaction can also be performed in strictly aqueous media, but the
solubility of phosphine 1 therein has a limit around 60 µM.
(13) This value is similar to that reported for the reaction of triphenylphosphine
with ethyl 2-azidoacetate in benzene (0.01 M-1 s-1). Leffler, J. E.; Temple,
R. D. J. Am. Chem. Soc. 1967, 89, 5235-5246.
0.015 M-1 s-1 13
. This value suggests that relatively high concen-
trations (i.e., 100 µM to mM) of compound 1 will be required to
label azides present at low levels within metabolically labeled
biomolecules at an appreciable rate. Therefore, the low back-
ground fluorescence of compound 1 will be particularly im-
portant. Efforts are underway to increase the rate of reaction
by modulating the substituents on phosphorus. The application of
this tool in the detection and quantitation of azide-labeled post-
translational modifications in vitro and in vivo is also of significant
future interest.
JA029013Y
9
J. AM. CHEM. SOC. VOL. 125, NO. 16, 2003 4709