A genetically encoded fluorescent amino acid

Article abstract of DOI:10.1021/ja062666k

The fluorescent amino acid l-(7-hydroxycoumarin-4-yl) ethylglycine 1 has been genetically encoded in E. coli in response to the amber TAG codon. Because of its high fluorescence quantum yield, relatively large Stoke's shift, and sensitivity to both pH and

Full text of DOI:10.1021/ja062666k

Published on Web 06/20/2006
A Genetically Encoded Fluorescent Amino Acid
Jiangyun Wang, Jianming Xie, and Peter G. Schultz*
Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037
Received April 17, 2006; E-mail: schultz@scripps.edu
The ability to selectively modify proteins with fluorescent probes
has greatly facilitated both in vitro and in vivo studies of protein
structure and function.1,2 For example, green fluorescent protein
(GFP) is a powerful probe of protein expression, localization, and
bimolecular interactions. However, GFP fusions are limited to the
C- or N- terminus of the target protein, relatively insensitive to
environment, and GFP can cause significant perturbations due to
its size.2 Chemical methods also can be used to selectively modify
proteins with a variety of synthetic fluorophores1,3 but are generally
limited to uniquely reactive surface accessible residues on isolated
proteins (e.g., the modification of cysteine with maleimide deriva-
tives).1 Biosynthetic labeling methods using chemically misacylated
tRNAs4 afford limited yields of protein and are typically carried
out in vitro. Methodology that would allow one to genetically
encode fluorescent amino acids at defined sites in proteins directly
in living organisms would significantly extend the scope of this
technique.5 Here we report the generation of an orthogonal tRNA/
aminoacyl-tRNA synthetase pair that allows the selective introduc-
tion of the fluorescent amino acid L-(7-hydroxycoumarin-4-yl)
ethylglycine 1 (Figure 1A) into proteins in E. coli in response to
the amber stop codon, TAG.
Figure 1. (A) Synthesis of L-(7-hydroxycoumarin-4-yl) ethylglycine 1. (a)
N,N′-Carbonyldiimidazole, rt, 2 h; (b) ethyl magnesium malonate, rt,
overnight; (c) resorcinol, methanesulfonic acid, rt, 2 h. (B) Absorption and
emission spectra of 1. Both spectra were recorded at pH 7.4, 100 mM sodium
phosphate buffer. At this pH, 1 is present in both phenol and phenolate
forms (in approximately a 2:1 ratio). The phenolate form has an extinction
coefficient of 17 000 at 360 nm and a quantum yield of 0.63.
The 7-hydroxycoumarin moiety was initially investigated due
to its high fluorescence quantum yield, relatively large Stoke’s shift
(Figure 1B), small size, and sensitivity to pH (Figure S1) and solvent
polarity.6 The coumarinyl amino acid 1 was synthesized by first
converting N-R-Cbz-L-glutamic acid R-benzyl ester into the side-
chain â-keto ester, which was then reacted with resorcinol in
methanesulfonic acid (von Pechmann reaction)7 to afford amino
acid 1 (Figure 1A). To selectively incorporate amino acid 1 at
defined sites in proteins in E. coli, a mutant Methanococcus
jannaschii tyrosyl amber suppressor tRNA (MjtRNATyrCUA)/tyrosyl-
tRNA synthetase (MjTyrRS) pair was evolved that uniquely
specifies 1 in response to the TAG codon.5
Three rounds of positive and two rounds of negative selections
of this library generated a clone that grows at 100 µg/mL
chloramphenicol in the presence of 1, but only at 20 µg/mL
chloramphenicol in the absence of 1. This clone CouRS-D8 has
the following eight mutations: Tyr32Glu, Leu65His, Ala67Gly,
His70Gly, Phe108Tyr, Gln109His, Asp158Gly, and Leu162Gly.
Four residues are mutated to glycine, most likely to create enough
space to accommodate 1. Tyr32 and Asp158 which hydrogen bond
to bound tyrosine in the wild-type enzyme are also mutated,
consistent with the loss of activity toward tyrosine.8,9
To determine the efficiency and fidelity for the incorporation of
1 into proteins, an amber codon was substituted for Ser-4 in sperm
whale myoglobin containing a C-terminal His6 tag. Protein expres-
sion was carried out in the presence of the selected synthetase
(CouRS-D8) and MjtRNATyrCUA in E. coli grown in either minimal
media or 2YT supplemented with 1 mM of 1. As a negative control,
protein expression was also carried out in the absence of 1. Analysis
of the purified protein by SDS-PAGE showed that the full length
protein was expressed only in the presence of 1 (Figure 2). The
yield of the mutant myoglobin in either media is 2 mg/L. For
comparison, the yield of wild-type myoglobin under similar
conditions is 5 mg/L. ESI-mass spectrometric analyses of the mutant
myoglobin gave an observed average mass of 18 511 Da, in close
agreement with the calculated mass of 18 512 Da (Figure S2,
Supporting Information).
To accommodate the large coumarin side chain, an MjTyrRS
library pBK-lib5 was generated in which His70 was mutated to
Gly and Ala67 was fixed as either Ala or Gly to increase the active
site size. Six residues, Tyr-32, Leu-65, Phe-108, Gln-109, Asp-
158, and Leu-162, in close proximity to bound tyrosine8,9 were then
randomized. Subsequently, this library was subjected to rounds of
both positive and negative selections. In the positive selection, cell
survival is dependent on the suppression of an amber codon
introduced at a permissive site in the chloramphenicol acetyl
transferase (CAT) gene when cells cotransformed with pBK-lib and
MjtRNATyrCUA are grown in the presence of 1 mM unnatural amino
acid (UAA) and chloramphenicol.10 Positively selected clones are
then transformed into cells containing MjtRNATyr
and a gene
CUA
encoding the toxic barnase protein with three amber mutations
introduced at permissive sites. These cells are grown in the absence
of UAA to remove any clones that utilize endogenous amino acids
(negative selection).
9
8738
J. AM. CHEM. SOC. 2006, 128, 8738-8739
10.1021/ja062666k CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
induced unfolding curves of the wild type, Ser4 f 1 and His37 f
1 holomyoglobins, as monitored by circular dichroism, are virtually
identical, suggesting that introduction of coumarin 1 into either helix
A or helix C does not significantly perturb protein stability. At 2
M urea concentration, the fluorescence intensity of 1 at position 4
in holomyogloin increases 30% (and remains at this level from 2
to 5 M urea), suggesting that this region of the protein is disordered.9
In contrast, mutant myoglobin with 1 at residue 37 shows little
change in its fluorescence intensity at 2 M urea, but undergoes a
similar fluorescence increase at 3 M urea. Consistent with this result,
a previous NMR study has shown that when the urea concentration
is higher than 2.2 M, helix A and B are largely disordered, as shown
by the disappearance of short and medium range NOEs in this
region, whereas helices C, D, and F unfold later, when the urea
concentration is higher than 3.0 M.11 Thus, it appears that amino
acid 1 is a site-specific probe of protein conformational changes,
in contrast to circular dichroism, which reports global changes
averaged over the entire structure.
Figure 2. Coomassie-stained SDS-PAGE (left) of TAG4 mutant myoglobin
(indicated by black arrow) expression in the presence and absense of 1
mM 1. The right panel shows the fluorescence image of wild-type and TAG4
mutant myoglobin.
The sensitivity of 1 to solvent polarity and pH should make it a
useful probe for many biological studies, both in vitro and in vivo.7
For example, amino acid 1 can be used to monitor bimolecular
interactions or conformational changes in proteins or the topology
of membrane-bound proteins. In addition, because 1 has a pKa of
7.8, which can be systematically altered by substitution of the
coumarin ring,12 it should be a useful probe of organellar pH and
pH-dependent cellular processes. Moreover, in its excited state,
7-hydroxycoumarin is both a strong photoacid13 and may facilitate
the study of proton-transfer processes in proteins.
Acknowledgment. This manuscript is dedicated to Peter B.
Dervan on the occasion of his 60th birthday. We thank Dr. Q. Zhang
and Dr. J. C. Anderson for helpful discussions. We are grateful to
the DOE (ER46051) and the Skaggs Institute for Chemical Biology.
Supporting Information Available: Materials and methods (PDF).
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Hermanson, G. T. Bioconjugate Techniques; Academic Press: San Diego,
CA, 1996.
(2) Tsien, R. Y. Annu. ReV. Biochem. 1998, 67, 509-44.
(3) (a) Martin, B. R.; Giepmans, B. N. G.; Adams, S. R.; Tsien, R. Y. Nat.
Biotechnol. 2005, 23, 1308-14. (b) Keppler, A.; Gendreizig, S.; Grone-
meyer, T.; Pick, H.; Vogel, H.; Johnsson, K. 2003, Nat. Biotechnol. 21,
86-9. (c) Lin, C.-W.; Ting, A. Y. J. Am. Chem. Sci. 2006, 128, 4542-3.
(4) (a) Mendel, D.; Cornish, V. W.; Schultz, P. G. Annu. ReV. Biophys. Biomol.
Struct. 1995, 24, 435-62. (b) Hohsaka, T.; Muranaka, N.; Komiyama,
C.; Matsui, K.; Takaura, S.; Abe, R.; Murakami, H.; Sisido, M. FEBS
Lett. 2004, 560, 173-7.
Figure 3. (A) Structure of sperm whale myoglobin (pdb code 105M).
Residues Ser4 and His37 are shown as sticks. (B) Urea-induced unfolding
of the Ser4 f 1 and His37 f 1 mutant holomyoglobins, monitored by
both CD and the fluorescent intensity of 1. Both experiments were carried
out in 100 mM sodium phosphate, 300 mM NaCl, pH 7.4, with various
concentrations of urea, as indicated. In the prescence of 5 M urea, both
myoglobin mutants show a 30% increase in fluorescence signal compared
to the signal in the presence of 0 M urea. The “fraction folded” is calculated
by normalizing the fluorescence signal or molar ellipticity at 5 M urea to
the fully unfolded state.
(5) (a) Wang, L.; Schultz, P. G. Angew. Chem., Int. Ed. 2005, 44, 34-66.
(b) Wang, L.; Xie, J.; Schultz, P. G. Annu. ReV. Biophys. Biomol. Struct.
2006, 35, 225-49.
(6) Zinsli, P. E. J. Photochem. 1974, 3, 55-69.
(7) Brun, M. P.; Bischoff, L.; Garbay, C. Angew. Chem., Int. Ed. 2004, 43,
3432-6.
(8) Kobayashi, T.; Nureki, O.; Ishitani, R.; Yaremchuk, A.; Tukalo, M.;
Cusack, S.; Sakamoto, K.; Yokoyama, S. Nat. Struct. Biol. 2003, 10, 425-
32.
To demonstrate the utility of 1, it was used as a probe of the
urea dependent denaturation of holomyoglobin. Because coumarin
fluorescence is sensitive to solvent polarity, its fluorescence intensity
should correlate with local unfolding of the protein in close
proximity to 1. The myoglobin structure consists of eight helices
(A to H) connected by short loops and turns.11 Ser4 in helix A and
His37 in helix C (both residues are largely solvent exposed and do
not significantly interact with other nearby residues) were each
mutated to 1 (Figure 3A). As indicated in Figure 3B, the urea
(9) Zhang, Y.; Wang, L.; Schultz, P. G.; Wilson, I. A. Protein Sci. 2005, 14,
1340-9.
(10) Xie, J. M.; Schultz, P. G. Methods 2005, 36, 227-38.
(11) Delli Castelli, D.; Lovera, E.; Ascenzi, P.; Fasano, M. Protein Sci. 2002,
11, 2273-6.
(12) Sun, W. C.; Gee, K. R.; Haugland, R. P. Bioorg. Med. Chem. Lett. 1998,
8, 3107-10.
(13) Cohen, B.; Huppert, D. J. Phys. Chem. A 2001, 105, 7157-64.
JA062666K
9
J. AM. CHEM. SOC. VOL. 128, NO. 27, 2006 8739