Cell-permeable near-infrared fluorogenic substrates for imaging β-lactamase activity

Article abstract of DOI:10.1021/ja042829+

This communication describes a design of cell-permeable near-infrared fluorogenic substrates for imaging β-lactamase expression in living mammalian cells. This design is based on fluorescence energy transfer resonance and utilizes a peracetylated d-glucos

Full text of DOI:10.1021/ja042829+

Published on Web 03/03/2005
Cell-Permeable Near-Infrared Fluorogenic Substrates for Imaging
â-Lactamase Activity
Bengang Xing, Ashot Khanamiryan, and Jianghong Rao*
Department of Radiology, Biophysics, Bio-X, Cancer Biology, and Molecular Imaging Programs at Stanford,
Stanford UniVersity School of Medicine, 300 Pasteur DriVe, Stanford, California 94305-5344
Received November 29, 2004; E-mail: jrao@stanford.edu
Scheme 1. Structures of CNIR1, CNIR2, CNIR3, and CNIR4 and
TEM-1 â-lactamase (Bla), the monomeric isoform product of
their Hydrolysis by Bla
the ampicillin resistance gene (ampr), is a small enzyme (29 kDa)
that efficiently hydrolyzes â-lactam substrates. It does not exist in
eukaryotes but can be easily expressed in eukaryotic cells without
any noticeable toxicity,1-3 and it has become an attractive biosensor
for detecting biological processes and interactions in vivo, such as
monitoring the promoter/regulatory elements’ activity,4-6 constitu-
tive and inducible protein interactions,7-9 and ribozyme in vivo
splicing reactions.10,11
Several fluorogenic substrates for Bla have been reported,4,12 but
none work for infrared or near-infrared fluorescence imaging.
Infrared/near-infrared light is preferred in molecular imaging studies
of living subjects because its long wavelength causes less photo
damage to cells, produces less autofluorescence background, and
offers better sensitivity. Furthermore, its longer wavelength has
better tissue penetration and less light scattering than visible light,
therefore it is more suitable for living animal imaging.13 Common
near-infrared dyes, including the carbocyanine series of compounds,
are widely used for labeling proteins and nucleic acids, but their
poor membrane permeability due to high molecular weights and
design that Bla hydrolysis of the substrate can lead to activation of
multiple charges makes it difficult to apply them for imaging living
the fluorescence of Cy5. Further analysis of the hydrolysis kinetics
subjects. Methods that include using synthetic grafted copolymers
of CNIR1 by Bla in phosphate-buffered saline (PBS) at pH 7.1
and nanoparticles have been developed to assist their cellular
revealed the catalytic constant, kcat ) 0.8 ( 0.1 s-1, and Michaelis
delivery.14,15 Here, we report a design of a cell-permeable small
constant, Km ) 6.7 ( 1.5 µM (Figure 2). Its catalytic efficiency
near-infrared fluorogenic substrate for Bla and its application in
imaging gene expression in living mammalian cells. Availability
of such a new substrate should further facilitate the application of
Bla as a biological reporter in living cells and even in whole living
animals.
(kcat/Km) is 7.4 × 105 M-1 s-1. The value of kcat is smaller than
that of a previously reported fluorogenic substrate, CCF2 (kcat
)
29 s-1),4 despite the chemical similarity in their 3′-leaving groups,
which might be related to the much larger size of CNIR1 as a
substrate for Bla. CNIR1 is highly stable with the spontaneous
hydrolysis rate constant in PBS of ∼1.7 × 10-7 s-1, and the
enzymatic acceleration (kcat/kuncat) is ∼5 × 106. A concentration as
low as 190 fM Bla is readily detectable with CNIR1.
Our designed substrates are shown in Scheme 1. Cy5 emits
maximally at 670 nm when excited at 650 nm. A quenching group,
QSY21, is not fluorescent and has a wide absorption spectrum from
540 to 730 nm with a peak at 660 nm. Therefore, they were chosen
as a fluorescence resonance energy transfer (FRET) pair so emission
from Cy5 can be efficiently quenched by QSY21. Cy5 was tethered
to the 7-amino of the cephalosporin through a glycyl linkage, and
QSY21 was attached to the 3′-position via a linker, which contained
an amino thiophenol and a cysteine residue. As an excellent leaving
group, the amino thiophenol at the 3′-position would facilitate the
fragmentation after Bla hydrolysis, and the cysteine introduces a
coupling site for adding any potential new functionality. The
substrate should have little or no fluorescence due to the FRET
quenching effect. Bla hydrolysis will break the FRET quenching
by releasing QSY21 and, thus, rendering a fluorescent product
containing Cy5 (Scheme 1). Preparation and characterization of
CNIR1 are detailed in the Supporting Information.
Because of its hydrophilic and charged nature, CNIR1 is not
cell-permeable. It has been reported that D-glucosamine can carry
molecules into cells through the attachment to its 2-amino group.16
To test this strategy, we decided to link D-glucosamine to the
Indeed, CNIR1 itself was essentially nonfluorescent, but pro-
duced a highly fluorescent product with a 57-fold increase in the
emission intensity at the wavelength of 660 nm upon treatment
with purified Bla (Figure 1). This large increase confirmed our
Figure 1. Fluorescence emission of CNIR1 (100 nM in PBS) before (‚ ‚ ‚)
and after (s) addition of Bla (excitation: 620 nm).
9
4158
J. AM. CHEM. SOC. 2005, 127, 4158-4159
10.1021/ja042829+ CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. Double-reciprocal plot of CNIR1 hydrolyzed per enzyme
molecule per second (V) versus substrate concentration. Error bars indicate
standard errors. Refer to Supporting Information for details.
Figure 3. Fluorescence (a and c) and differential interference contrast (b
and d) images of wild-type (a and b) and Bla stably transfected (c and d)
C6 glioma cells loaded with CNIR4 (4 µM) in Hank’s balanced salts
solution for 1 h at 37 °C. Cy5 emission is displayed in red. Fluorescence
images are corrected with background subtraction. See Supporting Informa-
tion for the detailed conditions.
carboxylate of cysteine on CNIR1 to facilitate its membrane
penetration. A γ-amino butyric acid was inserted as a spacer to
minimize any potential steric interactions between the transporter
and the enzyme. We have successfully prepared CNIR2 from
D-glucosamine. CNIR2 can efficiently detect Bla in vitro with a
58-fold increase in the fluorescence emission at 660 nm after the
treatment of Bla (Figure S2C in Supporting Information), which
suggested that the introduction of D-glucosamine has little interfer-
ence with its activity.
Fund (to J.R.). We thank Otsuka Chemical Co., Ltd. for the
generous gift of the lactam precursor, ACLH. B.X. thanks Dr. S.
Hasegawa for his assistance with cell imaging experiments.
We then tested the applicability of CNIR2 in imaging Bla activity
in living cells, CNIR2 was able to enter into cells but with just
moderate efficiency (data not shown). However, to our surprise, a
derivative of CNIR2 containing fully acetylated D-glucosamine,
CNIR3 (Scheme 1), was found to be much more efficient.
Incubation of both wild-type C6 glioma cells (no Bla expression;
a negative control) and stably transfected C6 glioma cells with
constitutive expression of Bla in a 4 µM solution of CNIR3 for 1
h at room temperature revealed a clear difference; in wild-type cells,
little Cy5 fluorescence was observed, and stably transfected C6
glioma cells emitted bright fluorescence signals with a contrast of
up to 10-fold higher.17
The peracetylated D-glucosamine in CNIR3 is presumably not
a substrate for glucosamine/glucose transporters, so its entrance into
cells may have a path different from that of native D-glucosamine.18
When incubated at 4 °C, little fluorescence signal was observed in
the Bla stably transfected C6 glioma cells, which may suggest the
possible involvement of endocytosis in the uptake.
Supporting Information Available: Synthetic procedures and
characterizations of all four near-infrared substrates, and procedures
for determining kinetic parameters and imaging Bla activity in living
cells (16 pages, PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Bush, K.; Sykes, R. B. In Antimicrobial Drug Resistance; Bryan, L. E.,
Ed.; Academic Press: New York, 1984; pp 1-31.
(2) Matagne, A.; Lamotte-Brasseur, J.; Frere, J. M. Biochem. J. 1998, 330,
581.
(3) Phlippon, A.; Dusart, J.; Joris, B.; Frere, J. M. Cell Mol. Life Sci. 1998,
54, 341.
(4) Zlokarnik, G.; Negulescu, P. A.; Knapp, T. E.; Mere, L.; Burres, N.; Feng,
L.; Whitney, M.; Roemer, K.; Tsien, R. Y. Science 1998, 279, 84.
(5) Zlokarnik, G. Methods Enzymol. 2000, 326, 221.
(6) Whitney, M.; Rockenstein, E.; Cantin, G.; Knapp, T.; Zlokarnik, G.;
Sanders, P.; Durick, K.; Craig, F. F.; Negulescu, P. A. Nat. Biotechnol.
1998, 16, 1329.
(7) Galarneau, A.; Primeau, M.; Trudeau, L.-E.; Michnick, S. W. Nat.
Biotechnol. 2002, 20, 619.
Acetylation of D-glucosamine improves the membrane perme-
ability of the substrate, but decreases its solubility. We, therefore,
introduced two sulfonate groups to QSY21 of CNIR3 and prepared
CNIR4 (Scheme 1).19 CNIR4 has a much improved solubility in
aqueous media with the cell permeability still maintained. The
images of Bla stably transfected C6 glioma cells were similar to
that of CNIR3, but the uptake was more uniform (Figure 3). The
fluorescence signals appeared to preferentially localize in endosomes
and some nuclei. When incubated at 37 °C, the uptake was even
higher and there was less background signal from wild-type cells;
thus the contrast was further improved (Figure 3). The retention of
the fluorescence signal in Bla-expressed cells seemed not to be
compromised even with 1 h incubation at 37 °C.
(8) Wehrman, T.; Kleaveland, B.; Her, J.-H.; Balint, R. F.; Blau, H. M. Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 3469.
(9) Spotts, J. M.; Dolmetsch, R. E.; Greenberg, M. E. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 15142.
(10) Hasegawa, S.; Jackson, W. C.; Tsien, R. Y.; Rao, J. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 14892.
(11) Hasegawa, S.; Choi, J. W.; Rao, J. J. Am. Chem. Soc. 2004, 126, 7158.
(12) Gao, W.; Xing, B.; Tsien, R. Y.; Rao, J. J. Am. Chem. Soc. 2003, 125,
11146.
(13) Weissleder, R. Nat. Biotechnol. 2001, 19, 316.
(14) Weissleder, R.; Tung, C.-H.; Mahmood, U.; Bogdanov, A. Nat. Biotechnol.
1999, 17, 375.
(15) Josephson, L.; Kircher, M. F.; Mahmood, U.; Tang, Y.; Weissleder, R.
Bioconjugate Chem. 2002, 13, 554.
In summary, we report here a design of cell-permeable near-
infrared fluorogenic substrates for â-lactamase and have demon-
strated their applicability in imaging â-lactamase activity in living
mammalian cells. With both good excitability with near-infrared
light and cell permeability, this new type of fluorogenic probe holds
promises for imaging gene expression in living animals, which is
currently ongoing in our laboratory.
(16) (a) Yoshioka, K.; Takahashi, H.; Homma, T.; Saito, M.; Oh, K.-B.;
Nemoto, Y.; Matsuoka, H. Biochim. Biophys. Acta 1996, 1289, 5. (b)
Ye, Y.; Bloch, S.; Achilefu, S. J. Am. Chem. Soc. 2004, 126, 7740.
(17) Refer to Figure S5 in Supporting Information for details.
(18) For examples of cellular uptake of acetylated saccharides, see: (a) Sarkar,
A. K.; Fritz, T. A.; Taylor, W. H.; Esko, J. D. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 3323. (b) Zhang, Z.; Gildersleeve, J.; Yang, Y.-Y.; Xu,
R.; Loo, J. A.; Uryu, S.; Wong, C.-H.; Schultz, P. G. Science 2004, 303,
371.
(19) The disulfonated derivative of QSY21 was prepared using the method
described in a patent: Haugland, R. P.; Singer, V. L.; Yue, S. T. U.S.
Patent 6,399,392 B1, June 4, 2002.
Acknowledgment. This work was supported by a grant from
the National Institutes of Health (EB003803-01) and the Career
Award at the Scientific Interface from the Burroughs Wellcome
JA042829+
9
J. AM. CHEM. SOC. VOL. 127, NO. 12, 2005 4159