β-galactosidase fluorescence probe with improved cellular accumulation based on a spirocyclized rhodol scaffold

Article abstract of DOI:10.1021/ja204781t

We identified a rhodol bearing a hydroxymethyl group (HMDER) as a suitable scaffold for designing fluorescence probes for various hydrolases. HMDER shows strong fluorescence at physiological pH, but phenolic O-alkylation of HMDER results in a strong prefe

Full text of DOI:10.1021/ja204781t

COMMUNICATION
pubs.acs.org/JACS
β-Galactosidase Fluorescence Probe with Improved Cellular
Accumulation Based on a Spirocyclized Rhodol Scaffold
||
||
||
Mako Kamiya,†,‡, ,# Daisuke Asanuma,†,‡, ,# Erina Kuranaga,‡,^ Asuka Takeishi,‡ Masayo Sakabe,‡,
||
||
Masayuki Miura,‡, Tetsuo Nagano,*,‡, and Yasuteru Urano*,†,§
†Graduate School of Medicine and ‡Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033,
Japan
§Basic Research Program and CREST, Japan Science and Technology Agency, 3-5 Sanbancho, Chiyoda-ku, Tokyo 102-0075, Japan
^Laboratory for Histogenetic Dynamics, RIKEN Center for Developmental Biology, 2-2-3 Minatojima-minamimachi, Chuo-ku,
Kobe 650-0047, Japan
S Supporting Information
b
independent of esterase activity. For this purpose, we reexamined
ABSTRACT: We identified a rhodol bearing a hydroxy-
methyl group (HMDER) as a suitable scaffold for designing
fluorescence probes for various hydrolases. HMDER shows
strong fluorescence at physiological pH, but phenolic O-al-
kylation of HMDER results in a strong preference for the
spirocyclic form, which has weak fluorescence. As a proof of
concept, we utilized this finding to develop a new fluores-
cence probe for β-galactosidase. This probe has favorable
characteristics for imaging in biological samples: it has good
cellular permeability, and its hydrolysis product is well-
retained intracellularly. It could rapidly and clearly visualize
β-galactosidase activity in cultured cells and in Drosophila
melanogaster tissue, which has rarely been achieved with
previously reported fluorescence probes.
available core fluorophores in order to identify one that would be
more likely to favor intracellular accumulation. As a candidate
scaffold, we focused on rhodol, which is a hybrid structure of
fluorescein and rhodamine.9,10 Rhodol derivatives are known to
be highly fluorescent in aqueous solution and are more resistant
to photobleaching in comparison with fluorescein.9 They have
been successfully used as scaffolds for fluorescence probes to
detect hydrogen peroxide, zinc ion, peroxynitrite, and hypoxia.11À15
Furthermore, from the viewpoint of sensor design, rhodols have
favorable characteristics for our purpose: they contain a phenolic
hydroxy group into which a β-galactopyranoside group can be
incorporated, and more importantly, they are likely to be
accumulated in living cells.16 Thus, we decided to develop a
new fluorescence probe for β-galactosidase based on rhodol.
Among the series of rhodols, we focused on N,N-diethylrhodol
(DER), which we found to have a pKa of 5.7 (Figure S2); this
should allow stable measurements near physiological pH. In
order to control the fluorescence of DER-based probes, we first
considered employing the same strategy used for our previously
reported TG-βGal probe,7 namely, precise control of fluores-
cence by means of a photoinduced electron transfer (PeT)
mechanism. However, this strategy was unfortunately not sui-
table for DER-based hydrolase probes because DER shows only a
small alteration in reduction potential in conversions between
the phenolate form and the alkylated form (Table S1).
Thus, we focused on other mechanisms to control the fluores-
cence of the fluorophore. It has been reported that conversion of
the carboxylate group of rhodamine derivatives to a mercapto-
methyl or carbamoyl group stabilizes the spirocyclic structure,
and this phenomenon has been utilized to control the fluores-
cence before/after reaction with targets such as hypochlorous
acid17,18 and cations.19,20 It is also known that conversion to a
hydroxymethyl group induces spirocycle formation of dichloro-
fluorescein under very basic conditions but not in the physiolo-
gical pH range.21 Since little is known to date about the effect of a
hydroxymethyl group on the fluorescence properties of other
fluorophores, including rhodol derivatives, we initially prepared
hydroxymethyl-DER (HMDER) and examined the effect of the
scherichia coli β-galactosidase is a well-characterized enzyme
E
that is widely used as a reporter to examine transcriptional
regulation or to evaluate transfection efficiency.1,2 Several fluoro-
genic substrates for visualizing the activity of this enzyme in living
cells and tissues have been developed, but most suffer from a
variety of disadvantages. For example, fluorescein di-β-galacto-
pyranoside (FDG)3 can only be loaded into cells by hypotonic
shock4,5 or must be applied to fixed or permeabilized cells,6 as it
shows poor cellular permeability. In contrast, TG-βGal, our
previously reported fluorescence probe based on the Tokyo-
Green (TG) scaffold, can detect β-galactosidase activity in living
cells7 and proved to be useful for efficiently selecting clones
expressing β-galactosidase on 96-well plates (unpublished data).
However, when applied to cultured cells or tissues such as
brain slices, it could not provide clear fluorescence images of
β-galactosidase-expressing cells or regions, mainly because its
intracellularly generated fluorescent hydrolysis product tends to
diffuse across the cell membrane, resulting in a poorly defined
signal (Figure S1 in the Supporting Information). A conventional
solution to this problem would be to modify the probe to obtain
improved cellular accumulation by introducing an esterase-
reactive acetoxymethyl (AM) group.8 However, this strategy
cannot be employed in vivo or in living tissues because of
the ubiquitous presence of esterase activity in extracellular fluid.
Therefore, we aimed to develop an alternative approach that is
Received: May 25, 2011
Published: July 25, 2011
r
2011 American Chemical Society
12960
dx.doi.org/10.1021/ja204781t J. Am. Chem. Soc. 2011, 133, 12960–12963
|

Journal of the American Chemical Society
COMMUNICATION
Figure 2. Enzymatic reaction of HMDER-βGal with β-galactosidase.
Absorption and fluorescence spectra of a 1 μM solution of HMDER-
βGal in PBS (pH 7.4) before and after addition of β-galactosidase
(5 units) are shown. The excitation wavelength was 525 nm.
Figure 1. (aÀc) Structures of (a) HMDER, (b) HMDER-Me, and (c)
HMDER-βGal mainly present at various pH values. The highlighted
structures are the putative forms predominantly present at physiological
pH. (d) pH dependence of the absorbance of HMDER, HMDER-Me,
and HMDER-βGal.
Figure 3. Detection of β-galactosidase activity in living cells. Fluores-
cence images of HEK/lacZ cells after incubation with a 10 μM solution
of (a) HMDER-βGal or (b) TG-βGal for 30 min are shown. Scale bars
represent 50 μm.
hydroxymethyl group by comparing the pH profiles of absor-
bance and fluorescence of DER and HMDER. The pKa of the
phenolic hydroxy group of HMDER was determined to be 5.4,
which is close to that of DER (Figure 1a and Figures S2 and S3).
The decreased absorption above pH 10 could be explained by the
involvement of a nonfluorescent spirocyclic form. In this paper,
we define pKcycl as the pH value at which the absorbance of the
compound decreases to half the maximum absorbance as a result
of the spirocyclization. The pKcycl of HMDER was determined to
be 11.3 (Figure 1a and Figure S3). In contrast, a methyl ether
derivative of HMDER (HMDER-Me) has a pKcycl of 6.6
(Figure 1b and Figure S4). Interestingly, at the physiological
pH of ∼7.4, HMDER is present as a colored and fluorescent
open form, while HMDER-Me is present in an equilibrium
between a colorless nonfluorescent spirocyclic form (85%) and
a weakly fluorescent open form (15%) (Figure 1d and Table S2).
These results show that the hydroxymethyl group stabilizes the
spirocyclic form of HMDER-Me but not that of HMDER at
physiological pH. We considered that this remarkable difference
in the pKcycl values of HMDER and HMDER-Me, which might
be due to the difference in their reduction potentials (Table S1),
could be utilized to design a rhodol-based probe for β-galacto-
sidase. That is, incorporation of an enzyme-reactive moiety
at the phenolic hydroxy group of HMDER would result in
suppression of the background fluorescence via spirocycle
formation, while reaction with the enzyme would generate
fluorescent HMDER.
To test this idea, we synthesized HMDER-βGal by incorpor-
ating a β-galactopyranoside group into HMDER (Figure 1c). We
examined the pH profile and fluorescence properties of HMDER-
βGal and obtained a pKcycl value of 6.9, meaning that it would be
present in an equilibrium between the spirocyclic form (75%)
and the open form (25%) at pH 7.4, similar to the case of
HMDER-Me (Figure 1c,d and Figure S5). The spirocyclic form
has no absorption/fluorescence in the visible wavelength region,
and the minor open form shows only weak fluorescence (Φfl =
0.009). In contrast, HMDER is predominantly present in the
open form, which shows relatively strong fluorescence (Φfl =
0.141) (Table S2).
As a result of these effects, a fluorescence increase of up to 76-
fold upon reaction with β-galactosidase was observed (Figure 2
and Figure S6). We next examined whether this new probe could
visualize β-galactosidase activity in living cells. When HEK/lacZ
cells were incubated with a 10 μM solution of HMDER-βGal, we
observed a clear fluorescence increase within the cells (Figure 3a
and Figure S7a), while no obvious fluorescence increase was seen
with HEK cells, which do not express β-galactosidase (Figure S8a).
In comparison with the fluorescence signal obtained with
TG-βGal (Figure 3b and Figure S7b), the signal obtained with
HMDER-βGal was confined within the cells and the surrounding
12961
dx.doi.org/10.1021/ja204781t |J. Am. Chem. Soc. 2011, 133, 12960–12963

Journal of the American Chemical Society
COMMUNICATION
Since β-galactosidase is one of the most widely used reporter
enzymes in biological studies, the potential range of application
of this new probe would be enormous. Finally, it should be easy
to extend this simple but effective design strategy, which involves
functionalization of the phenolic hydroxy group of HMDER, to
develop fluorescent probes for a wide variety of target molecules,
suchas other glycosidases (e.g., sialidases/mannosidases), esterases
(Figure S1), and reactive oxygen species. Such probes would be
powerful tools for biological studies.
’ ASSOCIATED CONTENT
S
Supporting Information. Synthesis and characterization
b
of compounds, pH profiles, fluorescence imaging, cytotoxicity
assay, experimental details, and a movie showing β-galactosidase
activity in D. melanogaster tissues (AVI). This material is available
free of charge via the Internet at http://pubs.acs.org.
Figure 4. Fluorescence imaging of β-galactosidase activity in D. mela-
nogaster wing disks with HMDER-βGal. Wing disks expressing β-
galactosidase were incubated with HMDER-βGal (50 μM) in PBS(À)
for 2 min. (a) Bright-field image. (b) Fluorescence image. (c) Merged image
of (a) and (b). (d) Image after X-Gal staining. Scale bars represent 100 μm.
’ AUTHOR INFORMATION
Corresponding Author
tlong@mol.f.u-tokyo.ac.jp; uranokun@m.u-tokyo.ac.jp
signal was suppressed to a low level, enabling us to achieve high-
contrast imaging. We further examined the subcellular localiza-
tion of the fluorescence signal produced by HMDER-βGal and
found that the cleaved dye tends to accumulate mainly in the
endoplasmic reticulum and partially in the Golgi apparatus, while
the enzyme is expressed diffusely in the cytosol (Figures S9 and
S10). Furthermore, the values of logP (the logarithm of the
partition coefficient) calculated with the ALOGPS 2.1 program22
were 2.88 ( 0.85 for HMDER-βGal (spirocyclic form) and
1.76 ( 0.68 for TG-βGal, while the logD values at pH 7.4
determined by an HPLC method were 2.0 for HMDER-βGal
and 1.1 for TG-βGal (Table S3). In both cases, HMDER-βGal
showed a value 1 unit larger than that of TG-βGal, suggesting
that HMDER-βGal should have improved cellular permeability.
We also examined the cytotoxicity of HMDER-βGal using an
MTT assay and observed no cytotoxic effect at concentrations up
to 100 μM (Figure S11).
Author Contributions
#These authors contributed equally.
’ ACKNOWLEDGMENT
This work was financially supported by a grant to T.N. from
the Hoansha Foundation and by research grants from the
Ministry of Education, Culture, Sports, Science and Technology
of Japan (Grants 20117003 and 23249004 to Y.U., and 23113504
to M.K.).
’ REFERENCES
(1) Alam, J.; Cook, J. L. Anal. Biochem. 1990, 188, 245.
(2) Spergel, D. J.; Kr€uth, U.; Shimshek, D. R.; Sprengel, R.; Seeburg,
P. H. Prog. Neurobiol. 2001, 63, 673.
(3) Rotman, B.; Zderic, J. A.; Edelstein, M. Proc. Natl. Acad. Sci.
Finally, we tried to visualize β-galactosidase activity in tissues
of Drosophila melanogaster, which is one of the most commonly
used model organisms in genetic studies.23,24 We used wing discs
(a precursor structure of the wing) taken from third-instar larvae,
in which β-galactosidase (used as a marker enzyme) is expressed
only at the posterior region. Soon after the start of incubation of
the tissue with HMDER-βGal, a marked fluorescence increase
was seen only in the β-galactosidase-expressing region (Figure 4).
The rate of fluorescence increase was so rapid that an image
could be obtained within just a few minutes (see the movie in the
Supporting Information). The expression of β-galactosidase was
confirmed by conventional X-Gal staining, which showed that
the fluorescence signal was colocalized with the β-galactosidase
activity. In contrast, when wing discs not expressing β-galactosidase
were used, there was no clear fluorescence increase (Figure S12).
In conclusion, we have designed, synthesized, and evaluated
HMDER-βGal, a new fluorescence probe for β-galactosidase. To
our knowledge, this is the first demonstration that spirocycliza-
tion of a hydroxymethyl group can be effectively utilized to
control fluorescence emission before/after reaction with the
target enzyme at physiological pH. Furthermore, our use of the
rhodol scaffold resulted in increased cellular accumulation of
the probe, enabling us to perform clear in situ visualization of
β-galactosidase activity in living cells and D. melanogaster tissues.
U.S.A. 1963, 50, 1.
(4) Nolan, G. P.; Fiering, S.; Nicolas, J.-F.; Herzenberg, L. A. Proc.
Natl. Acad. Sci. U.S.A. 1988, 85, 2603.
(5) Fiering, S.; Roederer, M.; Nolan, G. P.; Micklem, D. R.; Parks,
D. R.; Herzenberg, L. A. Cytometry 1991, 12, 291.
(6) Rakhmanova, V. A.; MacDonald, R. C. Anal. Biochem. 1998,
257, 234.
(7) Urano, Y.; Kamiya, M.; Kanda, K.; Ueno, T.; Hirose, K.; Nagano,
T. J. Am. Chem. Soc. 2005, 127, 4888.
(8) Kamiya, M.; Kobayashi, H.; Hama, Y.; Koyama, Y.; Bernardo,
M.; Nagano, T.; Choyke, P. L.; Urano, Y. J. Am. Chem. Soc. 2007,
129, 3918.
(9) Whitaker, J. E.; Haugland, R. P.; Ryan, D.; Hewitt, P. C.; Haugland,
R. P.; Prendergast, F. G. Anal. Biochem. 1992, 207, 267.
(10) Peng, T.; Yang, D. Org. Lett. 2010, 12, 496.
(11) Dickinson, B. C.; Chang, C. J. J. Am. Chem. Soc. 2008, 130,
9638.
(12) Dickinson, B. C.; Huynh, C.; Chang, C. J. J. Am. Chem. Soc.
2010, 132, 5906.
(13) Komatsu, H.; Harada, H.; Tanabe, K.; Hiraoka, M.; Nishimoto,
S.-i. Med. Chem. Commun. 2010, 1, 50.
(14) Peng, T.; Yang, D. Org. Lett. 2010, 12, 4932.
(15) Tomat, E.; Lippard, S. J. Inorg. Chem. 2010, 49, 9113.
(16) Kamino, S.; Ichikawa, H.; Wada, S.-i.; Horio, Y.; Usami, Y.;
Yamaguchi, T.; Koda, T.; Harada, A.; Shimanuki, K.; Arimoto, M.; Doi,
M.; Fujita, Y. Bioorg. Med. Chem. Lett. 2008, 18, 4380.
12962
dx.doi.org/10.1021/ja204781t |J. Am. Chem. Soc. 2011, 133, 12960–12963

Journal of the American Chemical Society
COMMUNICATION
(17) Kenmoku, S.; Urano, Y.; Kojima, H.; Nagano, T. J. Am. Chem.
Soc. 2007, 129, 7313.
(18) Koide, Y.; Urano, Y.; Kenmoku, S.; Kojima, H.; Nagano, T.
J. Am. Chem. Soc. 2007, 129, 10324.
(19) Dujols, V.; Ford, F.; Czarnik, A. W. J. Am. Chem. Soc. 1997, 119,
7386.
(20) Huang, J.; Xu, Y.; Qian, X. J. Org. Chem. 2009, 74, 2167.
(21) Koide, K.; Song, F.; de Groh, E. D.; Garner, A. L.; Mitchell,
V. D.; Davidson, L. A.; Hukriede, N. A. ChemBioChem 2008, 9, 214.
(22) Tetko, I. V.; Tanchuk, V. Y. J. Chem. Inf. Comput. Sci. 2002, 42,
1136.
(23) Rubin, G. M. Science 1988, 240, 1453.
(24) Kuranaga, E.; Kanuka, H.; Igaki, T.; Sawamoto, K.; Ichijo, H.;
Okano, H.; Miura, M. Nat. Cell Biol. 2002, 4, 705.
12963
dx.doi.org/10.1021/ja204781t |J. Am. Chem. Soc. 2011, 133, 12960–12963