HaloTag protein-mediated site-specific conjugation of bioluminescent proteins to quantum dots

Article abstract of DOI:10.1002/anie.200601197

On the dot: A genetically engineered haloalkane dehalogenase was used to conjugate Renilla luciferase to quantum dots (see picture). The quantum dots can emit light through bioluminescence resonance energy transfer (BRET). This specific conjugation occurs upon simple mixing under mild conditions, and may be applied for specific in vivo labeling of proteins with quantum dots for imaging. (Figure Presented).

Full text of DOI:10.1002/anie.200601197

Communications
quantum dot conjugates are either assembled through non-
specific interactions or prepared through site-nonspecific
coupling reactions.[9–18] For example, for proteins, the con-
jugation typically involves random amide coupling with either
amino- or carboxylate-presenting quantum dots.[9,15,16] Non-
specific conjugation chemistry leads to chemical heterogene-
ity of synthesized conjugates, may compromise the protein
activity and even induce aggregations, and is not applicable to
specific labeling of target proteins in vivo.[19] Specific non-
covalent interactions between receptors and ligands, such as
carbohydrate–lectin and streptavidin–biotin, have been
applied to assemble quantum-dot complexes.[1,20–21] Herein,
we report a specific conjugation method that utilizes a
genetically engineered hydrolase to covalently immobilize a
bioluminescent protein at the quantum-dot surface. This
immobilized bioluminescent protein can efficiently produce
chemical energy to excite quantum dots through resonance
energy transfer.
Luminescence Marker
DOI: 10.1002/anie.200601197
Our method employs a commercially available, engi-
neered haloalkane dehalogenase, the HaloTag protein
(HTP).[22] The native enzyme is a monomeric protein (MW
ꢀ 33 KDa) that cleaves carbon halogen bonds in aliphatic
halogenated compounds.[22] Upon nucleophilic attack by the
chloroalkane to Asp106 in the enzyme, an ester bond is
formed between the HaloTag ligand and the protein
(Scheme 1). HTP contains a critical mutation in the catalytic
triad (His272 to Phe) so that the ester bond formed between
HTP and HaloTag ligand cannot be further hydrolyzed
(Scheme 1).[22] HaloTag ligands labeled with small organic
dyes, such as coumarin and fluorescein, have been developed
for in vivo labeling of target proteins.[22] Herein we apply this
technology for the specific conjugation of proteins to
quantum dots.
To take advantage of this specific protein–ligand inter-
action, quantum dots can be functionalized with HaloTag
ligands. A protein target can in turn be genetically fused to
HTP at either its N- or C- terminus. The resulting fusion
protein can then be conjugated to quantum dots through the
reaction between HaloTag ligands and HTP (Scheme 1).
To demonstrate the utility of this method for quantum dot
conjugation, we chose a bioluminescent protein, Renilla
luciferase, as our target. We have recently demonstrated
that when Renilla luciferase is conjugated to quantum dots,
bioluminescence resonance energy transfer (BRET) can take
place.[23] Such quantum dot conjugates can emit light without
light excitation and offer greatly improved sensitivity for in
vivo imaging. With Renilla luciferase as the target protein for
the conjugation, the conjugation reaction can be conveniently
evaluated from the BRET emission of the quantum dots—a
measure of both the conjugation chemistry and the function
of the conjugated luciferase.
HaloTag Protein-Mediated Site-Specific
Conjugation of Bioluminescent Proteins to
Quantum Dots**
Yan Zhang, Min-kyung So, Andreas M. Loening,
Hequan Yao, Sanjiv S. Gambhir, and Jianghong Rao*
Quantum dots are fluorescent semiconductor nanocrystals
that have attracted much attention as fluorescence imaging
probes owing to their unique optical properties such as high
quantum yield, high molar extinction coefficients, narrow
emission spectra, size-dependent tunable emission, and high
photostability.[1–8] To apply quantum dots to biological
detection and imaging applications, quantum dots have to
be conjugated to molecules (e.g. peptide ligands, carbohy-
drate, antibodies, small molecule ligands) that can specifically
recognize the biological target under study. Numerous
examples have been reported on the use of quantum dots
for both in vitro assays and in vivo imaging, but these
[*] Dr. Y. Zhang,[+] M.-K. So,[+] Dr. H. Yao, Prof. S. S. Gambhir, Prof. J. Rao
Molecular Imaging Program at Stanford
Department of Radiology
Stanford University School of Medicine
1201 Welch Road, Stanford, CA 94305-5484 (USA)
Fax: (+1)650-736-7925
E-mail: jrao@stanford.edu
A. M. Loening, Prof. S. S. Gambhir
Department of Bioengineering
Stanford University
Stanford, CA 94305 (USA)
[+] Y. Zhang and M.-K. So equally contributed to this work.
A stabilized mutant of Renilla luciferase (Luc8) was
genetically fused to the N terminus of the HTP and expressed
to obtain the fusion protein HTP–Luc8. The C terminus of
HTP–Luc8contained a 6 ” His tag to facilitate its purification.
Gel electrophoresis analysis indicated that the molecular
weight of the fusion protein was consistent with the expected
value, approximately 70 kDa (Figure 1a). The biolumines-
cence activity of the fusion protein was estimated to be 1.2 ”
[**] This work was supported by the Burroughs Wellcome Fund (to J.R.),
a Stanford School of Medicine Dean’s Fellowship (to Y.Z.), the Korea
Research Foundation Grant M07-2004-000-10234-0 (to M.K.S.), a
Stanford Bio-X Graduate Fellowship (to A.M.L.), and the National
Cancer Institute Centers of Cancer Nanotechnology Excellence
(CCNE) U54.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4936
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 4936 –4940

Angewandte
Chemie
Scheme 1. Schematic of the specific conjugation of proteins to quantum dots mediated by the HaloTag protein and its ligand. HTL=HaloTag
ligand.
1023 photonssÀ1 molÀ1, which is approximately 86% of Luc8
(Figure 1b).
To minimize potential steric hindrance between the
quantum dots and HaloTag proteins during conjugation, we
designed a HaloTag ligand containing an amino ethylene
glycol group that would help orient the ligand away from the
quantum-dot surface (Scheme 1). The HaloTag ligand 1 was
prepared from 6-chloro-1-iodohexane and 2-(2-aminoethox-
y)ethanol by the synthetic route outlined in Scheme 2, and
was then immobilized through its amino group to the
carboxylate-presenting quantum dots (QD@COOH). The
resulting quantum dots coated with the HaloTag ligand 1
(QD@1) showed good solubility in neutral pH buffer solution.
The conjugation of the fusion protein HTP–Luc8to QD@ 1
was carried out by a simple mixing of both at 378C, resulting
in the formation of an irreversible covalent bond between
ligand 1 on quantum dots and HTP–Luc8.
As successful immobilization of HTP–Luc8to quantum
dots should allow BRET to occur, we measured the BRET
emission from the quantum dots to follow the conjugation
reaction. Addition of coelenterazine, the substrate for Renilla
luciferase, to the purified conjugate QD@1–HTP–Luc8
resulted in a dual-peak bioluminescence emission spectrum
(Figure 2). In addition to the Luc8peak at 480 nm, there was
an emission maximum at 655 nm that overlapped well with
the fluorescence emission of the quantum-dot conjugates
excited at 480 nm.
To evaluate whether the observed BRET emission was
owing to specific conjugation between the quantum dots and
fusion proteins, we examined the dependence of the BRET
emission on the HaloTag ligand 1 that was used in the
conjugation. QD@COOH was first functionalized with var-
ious concentrations of 1. These modified quantum dots were
subsequently conjugated with 20 equivalents of HTP–Luc8.
Figure 3 shows that with increasing amounts of ligand 1 used
in the conjugation, the bioluminescence emissions from both
the immobilized HTP–Luc8and the quantum dots through
Figure 1. Characterization of the size and function of the fusion
protein. a) Gel electrophoresis analysis confirmed the size of HTP–
Luc8. Both proteins (0.5 mg each) were run on a 4–12% Bis–Tris
(bis(2-hydroxyethyl)amino–tris(hydroxymethyl)methane) gradient dena-
turing gel and stained with Coomassie Blue. The expected sizes for
Luc8 and HTP–Luc8 were 37.1 and 70.3 kDa, respectively. b) Biolu-
minescence emission spectra of Luc8 (solid line) and HTP–Luc8
(dashed line). The inset shows the total photon production of Luc8
and HTP–Luc8.
Angew. Chem. Int. Ed. 2006, 45, 4936 –4940
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4937

Communications
number of immobilized proteins if the
BRET distance remains unchanged.[23]
This small decrease in the BRET ratio
may be due to a shift in the orientation of
the conjugated fusion proteins on the
quantum dot surface.[25]
Finally, we examined the dependence of
the conjugation reaction on the amount of
fusion-protein present. Quantum dots were
reacted with 1000 equivalents of HaloTag
ligand 1, and then reacted with increasing
concentrations of HTP–Luc8. As expected,
the resulting conjugates showed increasing
bioluminescence emissions both from
HTP–Luc8and from the quantum dots
(Figure 4). As a control, Luc8(without
HTP fusion) was incubated with the QD@
1. The control reaction showed no biolu-
Scheme 2. Synthesis of HaloTag ligand 1. Reagent and conditions: a) Boc2O/EtOH, 08C, 2 h;
b) NaH/DMF-THF and 6-chloro-1-iodohexane; c) TFA/anisole in DCM; K2CO3/MeOH; d) NaH/DMF-
THF and iodoacetic acid sodium salt; e) N-Hydroxysuccinimide and DCC in DCM; f) DIPEA/THF;
g) TFA/anisole in DCM; K2CO3/MeOH. Boc=tert-butoxycarbonyl, DMF=N,N-dimethylformamide,
TFA=trifluoroacetic acid, DCM=dichloromethane, DCC=1,3-dicyclohexylcarbodiimide, DIPEA=
diisopropylethylamine.
Figure 2. Bioluminescence (dashed line) and fluorescence (black solid
line) spectra of conjugate QD@1–HTP–Luc8 in borate buffer solution.
The fluorescence emission was collected with excitation at 480 nm.
BRET increased. When quantum dots (without a HaloTag
ligand 1 attached) were similarly mixed with the fusion
protein, there was only a small emission from HTP–Luc8at
480 nm and a small BRET emission from the quantum dots.
The small BRET emission probably arises from an electro-
static interaction between the 6 ” His tag on HTP–Luc8and
the negative carboxylate groups on the quantum dots. These
results confirm that the BRET emission reflects specific
conjugation occurring between the quantum dots and the
fusion protein HTP–Luc8, and that the fusion protein HTP–
Luc8. retains its enzymatic activity after conjugation.
The efficiency of the resonance energy transfer process
can be quantitatively estimated from the BRET ratio. The
BRETratio is defined by the acceptor emission relative to the
donor emission.[23,24] In the quantum dot and HTP–Luc8
conjugate, the donor is Luc8and the acceptor is the quantum
dot. We calculated the BRET ratio by dividing the total
emission from quantum dots by the total emission from HTP–
Luc8, shown in Figure 3b. With the increase in the number of
HaloTag ligand 1 and in turn the increase in immobilized
fusion protein, the BRET ratio decreased gradually from 0.6
to 0.4. In principle, the BRET ratio should not depend on the
Figure 3. Dependence of the conjugation on the ligand 1. a) Represen-
tative bioluminescence emission spectra of the conjugates synthesized
at different concentrations of 1 (from bottom to top: 0, 20, 100, 500,
1000, and 2500 equivalents). b) Total bioluminescence emissions from
HTP–Luc8 and from quantum dots, and the calculated BRET ratios of
conjugates prepared in (a; in duplicate).
4938 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 4936 –4940

Angewandte
Chemie
Synthesis of 1: the compound was synthesized from 6-chloro-1-
iodohexane and 2-(2-aminoethoxy)ethanol according to Scheme 2.
1H NMR (400 MHz, CDCl3): d = 3.92 (s, 2H), 3.70–3.20 (m, 18H),
2.09 (m, 2H), 2.64 (m, 2H), 2.50 (m, 2H), 1.40–1.20 ppm (s, 4H); LC-
MS: m/z 369.2 [M+1]+; calcd M+: 368.2.
Conjugation of 1 to quantum dots: Quantum dots, HaloTag ligand
1, and EDC (400 equiv) were mixed together in borate buffer solution
(10 mm, pH 7.4) and incubated at room temperature for 1 h. QD@1
was separated from free HaloTag ligand and excess EDC by filtration
through a 100 K NanoSep filter. The quantum dot conjugates were
washed three times with pH 8.5 borate buffer solution for 1 h before
being recovered with pH 7.4 borate buffer solution. The concentra-
tion of QD@1 was determined from the fluorescence intensity.
Preparation and purification of HTP–Luc8: the plasmid pBAD–
Luc8–HaloTag encoded for the fusion protein was constructed from
plasmid pBAD–RLuc8and plasmid pHT2 by PCR and ligation. E.
coli LMG194 cells transformed with this plasmid were induced with
0.2% arabinose and grown at 328C to an optical density at a
wavelength of 600 nm (OD600) of 0.7. Cells were lysed by thawing in
wash buffer solution (WB; NaCl (300 mm), 2-[4-(2-hydroxyethyl)-1-
Figure 4. Bioluminescence emission spectra of quantum dots conju-
gated with HTP–Luc8 (going down from the top line: 100, 50, 20, and
10 equivalents) or 20 equivalents of Luc8 (bottom dash–dot line). The
quantum dots were reacted with 1000 equivalents of HaloTag ligand 1
before the conjugation with HTP–Luc8. Unconjugated proteins were
removed by filtration before measurement.
piperazinyl]ethanesulfonic acid
( HEPES; 20 mm), imidazole
(20 mm), pH 8) containing lysozyme (1 mgmLÀ1), RNAse A
(10 mgmLÀ1), and DNAse I (5 mg/mL). Lysates were clarified by
centrifugation and allowed to bind to nickel affinity resin (Ni-NTA
Superflow, Qiagen) for 1 h at 48C with gentle mixing. After washing
with WB, the protein was eluted with elution buffer solution (NaCl
(300 mm), HEPES (20 mm), imidazole (250 mm), pH 8) and further
purified by anion-exchange chromatography (Source 15Q resin, GE/
Amersham) followed by gel-filtration chromatography with borate
buffer solution.
minescence emission, therefore indicating no immobilization
of Luc8on the quantum dots and further confirming that the
conjugation between quantum dots and HTP–Luc8was
specific.
In our previous demonstration of self-illuminating quan-
tum dots for in vivo imaging, the quantum dots were
conjugated with Luc8in vitro before their introduction into
living cells and animals. An important further step will be to
specifically conjugate luciferases with quantum dots for
functional imaging in vivo. The mild conjugation conditions
used to immobilize proteins to quantum dots, mediated by the
HaloTag protein and its ligand, may allow this method to be
applied to specific labeling of target proteins with quantum
dots in vivo. This method also offers an advantage in
comparison to a widely used conjugation method based on
biotin and streptavidin in that the HaloTag protein is
monomeric and relatively small.
Conjugation of QD@1 with HTP–Luc8: Typically QD@1
(5 pmol) was incubated with of HTP–Luc8(20 equiv) in borate
buffer solution (10 mm, pH 7.4) at 378C for 30 min. Free HTP–Luc8
was removed from the incubation mixture by filtration through a
100 K NanoSep filter at 48C. The filtered conjugates were washed
efficiently with pH 7.4 borate buffer solution at 48C. The final
quantum dot conjugates were recovered with ice-cold pH 7.4 borate
buffer solution.
Fluorescence and bioluminescence spectra were collected with a
Fluoro Max-3 (Jobin Yvon Inc.). Bioluminescence spectra were
acquired with the excitation light blocked.
Received: March 26, 2006
Revised: May 19, 2006
Published online: June 29, 2006
In summary, this communication reports a new method,
based on the specific interaction between the HaloTag protein
and its ligand, to functionalize quantum dots for biological
imaging. By using this method, we successfully conjugated a
bioluminescent protein to quantum dots and produced self-
illuminating quantum dot conjugates. This specific conjuga-
tion under mild physiological conditions offers promises for
specific in vivo labeling of proteins or cells with quantum dots
for imaging.
Keywords: bioluminescence resonance energy transfer ·
conjugation · luciferase · protein structures · quantum dots
.
[1] M. Bruchez, M. Moronne, P. Gin, S. Weiss, A. P. Alivisatos,
Science 1998, 281, 2013 – 2016.
[2] W. C. W. Chan, S. Nie, Science 1998, 281, 2016 – 2018.
[3] I. L. Medintz, H. T. Uyeda, E. R. Goldman, H. Mattoussi, Nat.
Mater. 2005, 4, 435 – 446.
[4] X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose,
J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir, S. Weiss,
Science 2005, 307, 538– 544.
[5] C. M. Niemeyer, Angew. Chem. 2001, 113, 4254 – 4287; Angew.
Chem. Int. Ed. 2001, 40, 4128– 4158.
[6] J. K. Jaiswal, S. M. Simon, Trends Cell Biol. 2004, 14, 497 – 504.
[7] B. O. Dabbousi, J. Rodriguez-Viejo, F. V. Mikulec, J. R. Heine,
H. Mattoussi, R. Ober, K. F. Jensen, M. G. Bawendi, J. Phys.
Chem. B 1997, 101, 9463 – 9475.
[8] C. A. Leatherdale, W. K. Woo, F. V. Mikulec, M. G. Bawendi, J.
Phys. Chem. B 2002, 106, 7619 – 7622.
Experimental Section
Chemicals for HaloTag ligand synthesis were from Sigma-Aldrich.
The coupling reagent 1-ethyl-3-(3-dimethylaminopropyl)carbodii-
mide hydrochloride (EDC) was from Fluka. Quantum dots were
from Invitrogen and have typical CdSe/ZnS core-shell structures with
the quantum yield (determined in 50 mm pH 9 borate buffer solution)
of 83%. Coelenterazine was from Prolume. The plasmid pHT2
(HaloTag) was from Promega. NanoSep 100 K filters for quantum dot
purification were from Pall, Life Science.
Angew. Chem. Int. Ed. 2006, 45, 4936 –4940
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4939

Communications
[9] S. Kim, Y. T. Lim, E. G. Soltesz, A. M. De Grand, J. Lee, A.
Nakayama, J. A. Parker, T. Mihaljevic, R. G. Laurence, D. M.
Dor, L. H. Cohn, M. G. Bawendi, J. V. Frangioni, Nat. Biotech-
nol. 2004, 22, 93 – 97.
[10] R. Charkrabarti, A. M. Klibanov, J. Am. Chem. Soc. 2003, 125,
12531 – 12540.
[11] M. E. Akerman, W. C. W. Chan, P. Laakkonen, S. N. Bhatia, E.
Ruoslahti, Proc. Natl. Acad. Sci. USA 2002, 99, 12617 – 12621.
[12] H. Mattoussi, J. M. Mauro, E. R. Goldman, G. P. Anderson, V. C.
Sundar, F. V. Mikulec, M. G. Bawendi, J. Am. Chem. Soc. 2000,
122, 12142 – 12150.
[13] S. J. Rosenthal, I. Tomlinson, E. M. Adkins, S. Schroeter, S.
Adams, L. Swafford, J. McBride, Y. Wang, L. J. DeFelice, R. D.
Blakely, J. Am. Chem. Soc. 2002, 124, 4586 – 4594.
[14] X. Gao, Y. Cui, R. M. Levenson, L. W. K. Chung, S. Nie, Nat.
Biotechnol. 2004, 22, 969 – 976.
[15] F. Pinaud, D. King, H.-P. Moore, S. Weiss, J. Am. Chem. Soc.
2004, 126, 6115 – 6123.
[16] X. Wu, H. Liu, J. Liu, K. N. Haley, J. A. Treadway, J. P. Larson, N.
Ge, F. Peale, M. P. Bruchez, Nat. Biotechnol. 2003, 21, 41 – 46.
[17] I. L. Medinitz, A. R. Clapp, H. Mattoussi, E. R. Goldman, B.
Fisher, J. M. Mauro, Nat. Mater. 2003, 2, 630 – 638.
[18] H. Mattoussi, J. M. Mauro, E. R. Goldman, T. M. Green, G. P.
Anderson, V. C. Sundar, M. G. Bawendi, Phys. Status Solidi B
2001, 224, 277 – 283.
[19] For example, conjugation of biotin to firefly luciferase via non-
specific N-hydroxysuccinimide ester coupling significantly
decreases luciferase activity. See H. Arakawa, M. Maeda, A.
Tsuji, Anal. Lett. 1992, 25, 1055 – 1063.
[20] X.-L. Sun, W. Cui, C. Haller, E. L. Chaikof, ChemBioChem 2004,
5, 1593 – 1596.
[21] M. Howarth, K. Takao, Y. Hayashi, A. Y. Ting, Proc. Natl. Acad.
Sci. USA 2005, 102, 7583 – 7588.
[22] Technical Manual, Promega, “HaloTag Interchangeable Label-
ing Technology”; http://www.promega.com, 2006.
[23] M.-K. So, C. Xu, A. M. Loening, S. S. Gambhir, J. Rao, Nat.
Biotechnol. 2006, 24, 339 – 343.
[24] K. D. G. Pfleger, K. A. Eidne, Nat. Methods 2006, 3, 165 – 174.
[25] A larger number of immobilized proteins would lead to denser
packing at the surface, resulting in the Luc8domain of the fusion
proteins pushed further away from the quantum dot surface and
consequently a larger average BRET distance.
4940 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 4936 –4940