Activatable rotor for quantifying lysosomal viscosity in living cells

Article abstract of DOI:10.1021/ja311688g

We have developed Lyso-V, the first fluorescent probe of lysosomal viscosity. Because of its lysosome-actived fluorescence characteristics, Lyso-V has proved to be an ideal lysosomal tracer with high spatial and temporal resolution under laser confocal mi

Full text of DOI:10.1021/ja311688g

Communication
pubs.acs.org/JACS
Activatable Rotor for Quantifying Lysosomal Viscosity in Living Cells
Lu Wang,† Yi Xiao,*,† Wenming Tian,‡ and Liezheng Deng‡
†State Key Laboratory of Fine Chemicals, Dalian University of Technology, Dalian 116024, China
‡State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian
116023, China
S
* Supporting Information
To our knowledge, fluorescent probes, which are powerful
ABSTRACT: We have developed Lyso-V, the first
fluorescent probe of lysosomal viscosity. Because of its
lysosome-actived fluorescence characteristics, Lyso-V has
proved to be an ideal lysosomal tracer with high spatial and
temporal resolution under laser confocal microscopy.
More importantly, Lyso-V shows its practical applicability
in real-time quantification of lysosomal viscosity changes in
live cells through fluorescence lifetime imaging micros-
copy.
tools in cell biology, can hardly meet the requirements for
monitoring lysosomal viscosity. Recently, several research
teams, including the groups of Haidekker,8 Suhling,9 and
Peng,10 have developed a number of viscosity-sensitive probes
based on different kinds of molecular rotors that respond to the
increased viscosity by enhancement of the fluorescence
intensity or increases in the fluorescence lifetime. These
inspiring works have enriched our knowledge of molecular
design ideas for viscosity probes. However, almost all of these
probes showed uneven, uncharted, and nonspecific intracellular
distributions that might influence the reliability of their
viscosity mapping. The only two exceptions were reported by
Haidekker8b and Suhling,9 who developed molecular rotors
containing long chains with high lipophilicity to bind to
membranes or other hydrophobic domains of cellular
organelles and to probe the local microviscosity of these
regions. To date, none of these fluorescent probes is well-suited
for lysosomal applications.
etection of intracellular viscosity remains one of the
D
challenges in chemical biology, although for a long time
viscosity has been well-recognized as an essential micro-
environmental parameter in a variety of biological systems.
Viscosity contributes to biological functions by affecting the
mobility of substances.1 Also, its abnormal changes are closely
linked with disorders and diseases (e.g., diabetes, infarction, and
hypertension).2 However, conventional viscometers suitable for
macroscopic fluids cannot be applied in live cells, necessitating
the development of new tools.3 One predictable obstacle is the
fact that the intracellular microenvironment, which is composed
of many different organelles (e.g., mitochondria and
lysosomes), is highly inhomogeneous. Not surprisingly, the
local viscosity differs considerably from one region to another.
For the sake of better indicating cells’ local changes, sensing the
intracellular viscosity (and other environmental parameters,
e.g., pH and polarity, and important chemical species, e.g., Ca2+
and nitric oxide) should to the greatest possible extent be
performed within distinct organelles rather than in an unknown
intracellular area.4
In this work, we developed Lyso-V, the first lysosome
molecular probe for viscosity, by attaching a morpholine moiety
to a typical boron dipyrromethene (BODIPY) molecular rotor
(Scheme 1). The rotor unit provides sensitivity toward the
Scheme 1. Working Principle of Lysosomal Viscosity Probe
Lyso-V
Lysosomal viscosity reflects the status and function of this
kind of organelle, whose locations, morphologies, and
components are always changing. Normally, as the cell’s
recycling centers, lysosomes receive senescent cellular debris
or needless macromolecules and process them into small
molecules that the cell can utilize.5 However, in cases of
lysosomal dysfunction, especially lysosomal storage diseases6
caused by the deficiency of single lysosomal enzymes,
macromolecular substances do not decompose but accumulate
within lysosomes. Since the lysosomal viscosity fluctuates along
with changes in the amounts and densities of such macro-
molecules, real-time monitoring of its dynamic alteration would
be very meaningful for not only fundamental cell biology but
also diagnosis of lysosome-related disorders.7 Unfortunately,
detecting lysosomal viscosity in living cells remains a “blank
space”, without any attempt reported in the literature.
viscosity because of the free rotation around the single bond
connecting the BODIPY core and the phenyl group.9a The
rotor can freely rotate, without being bound to proteins [see
the Supporting Information (SI)]. BODIPYs themselves are
highly lipophilic, giving them a large tendency to accumulate in
lipid-rich intracellular regions rather than in lysosomes. Within
Lyso-V, moderate alkalinity of the morpholine moiety is critical
for its specific localization in acidic lysosomes. According to our
previous study,11 morpholine has a pKa of 5−6, so it can be
protonated only in lysosomes (pH 4.5−5.5) but not in cytosol
Received: November 30, 2012
Published: February 14, 2013
© 2013 American Chemical Society
2903
dx.doi.org/10.1021/ja311688g | J. Am. Chem. Soc. 2013, 135, 2903−2906

Journal of the American Chemical Society
Communication
or in other subcellular organelles. Once Lyso-V diffuses into
lysosomes and is protonated, it becomes more hydrophilic and
is retained in the lysosomes. In addition, introduction of the
morpholine unit also endues Lyso-V with a lysosome-actived
fluorescence characteristic that is very beneficial for lysosomal
tracing. This occurs because the electron-rich morpholine unit
quenches the BODIPY fluorescence through photoinduced
electron transfer (PET).12 Thus, even though there is a trace
amount Lyso-V distributed outside the lysosomes, its back-
ground fluorescence is negligible. However, in lysosomes, the
morpholin moiety is protonated and can no longer serve as an
electron donor for PET. Hence, only Lyso-V molecules that are
localized in lysosomes are fluorescent.
Lyso-V’s fluorescence intensity is highly sensitive to both pH
and viscosity (Figure 1). On the one hand, at a particular
Figure 2. (a, b) Fluorescence lifetime spectra at 517 nm for 2 μM (a)
Lyso-V and (b) protonated Lyso-V [obtained by acidifying methanol/
glycerol systems with 4 μL of 37% HCl(aq)] in methanol/glycerol
mixtures with varying proportions to adjust the viscosity. (c)
Fluorescence lifetime vs viscosity curves for neutral (black) and
protonated (red) Lyso-V. (d) Fluorescence lifetimes of Lyso-V at
various pH in water/methanol/glycerol mixtures with varying portions
to adjust the viscosity to the four indicated levels and diluted with HCl
and NaOH solutions to adjust the pH.
Figure 1. (a, b) Fluorescence spectra of Lyso-V at different pH in
methanol/water/glycerol mixtures at two viscosities: (a) 143 and (b)
48 cP. (c) Fluorescence quantum yields of Lyso-V at different pH.
quenching” by the very fast PET process due to the short
distance between the fluorophore and the quencher. It was also
found that at different viscosities, the fluorescence lifetimes at
the same pH were quite different. When we precisely tuned and
determined the pH, we had to add large portions of water (see
the SI), and the as-obtained water/methanol/glycerol ternary
systems would have had considerably different polarities than
the binary methanol/glycerol mixtures, as water has a much
higher dielectric constant (ε = 78) than methanol (ε = 33) and
glycerol (ε = 43). However, it is worth noting that the lifetimes
determined in the two viscosity buffer systems were coincident.
The lifetimes at 69, 30, 10, and 0.7 cP were around 2.10, 1.43,
1.02, and 0.25 ns, respectively, over a range of pH (Figure 2d);
at all four viscosities, we found the identical coordinate
positions in the viscosity−lifetime curves for neutral and
protonated Lyso-V (Figure 2c). These results indicate that
Lyso-V’s fluorescence lifetime is sensitive only to the viscosity
and is independent of changes in pH and polarity. Hence, the
viscosity−lifetime curve in Figure 2c can be used as a
calibration curve for quantification of the viscosity by
fluorescence lifetime imaging microscopy (FLIM).14
The cytotoxicity of Lyso-V is very low within 24 h of
incubation (see the SI). An MTT experiment showed that
>90% of MCF-7 cells survived after 12 h (5 μM Lyso-V
incubation), and after 24 h the cell viability remained at ∼80%.
For many known lysosomal probes, long incubation can induce
an increase in the cell death rate. Compared with those of most
commercial lysosomal probes, the cytotoxicity of Lyso-V
toward MCF-7 cells is very low, which is a favorable property
for applications in living cells.
viscosity (e.g., 143 or 48 cP), Lyso-V’s fluorescence band at 517
nm increased as the pH decreased from 6 to 4; for example, at a
viscosity of 48 cP, the fluorescence quantum yield (ϕ) at pH 4
was 0.21, which is almost 20 times that at pH 7 (ϕ = 0.011).
On the other hand, when the pH was kept the same, the
fluorescence intensity decreased with decreasing viscosity. For
instance, at pH 4, ϕ was 0.21 at 48 cP, which is only two-thirds
of that at 143 cP (ϕ = 0.33). The above discovery revealed that
Lyso-V’s fluorescence intensity is not suitable for quantifying
viscosity because of the interference from pH changes. Because
the pH of lysosomes varies from 4 to 6 at different stages, laser
confocal imaging (fluorescence intensity imaging) might not be
a good choice for accurately determining the lysosomal
viscosity. However, it should be noted that for pH > 7, Lyso-
V’s fluorescence became very weak and negligible at all
viscosities. This characteristic, ascribed to PET from the neutral
morpholine donor to the excited BODIPY, helps eliminate the
background fluorescence outside the lysosomes and thus is very
beneficial for the use of Lyso-V to trace the locations of
lysosomes under laser confocal microscopy.
The fluorescence lifetime of Lyso-V exhibits satisfactory
specificity and sensitivity toward viscosity changes. First, the
fluorescence lifetimes of both Lyso-V and its protonated form
(obtained by acidifying the medium with a small but sufficient
amount of HCl) were measured in a series of viscosity gradient
buffers composed of methanol and glycerol in various
proportions (Figure 2a,b). Interestingly, the lifetime−viscosity
curve of neutral Lyso-V coincides with that of its protonated
counterpart, as they displayed almost identical lifetimes at every
viscosity (Figure 2c). Along with the gradual increase in
viscosity from 0.6 to 359.6 cP, the liftetimes became remarkably
longer, from 0.30 to 4.62 ns. Second, at the four chosen
viscosity levels (0.7, 10, 30, and 69 cP), the fluorescence
lifetimes were determined at different pH. At a given viscosity
level, the lifetimes at different pH were very similar; according
to Sauer and co-workers,13 this can be ascribed to “static
Lyso-V displayed strong localized fluorescence within
lysosomes. MCF-7 cells were costained with Lyso-V and
Neutral Red (NR), a commercial lysosome tracker, for 3 min at
37 °C. Lyso-V showed bright-green fluorescence in channel 1
(Figure 3b). At the same time, NR showed red fluorescence in
channel 2 (Figure 3c). The merged image (Figure 3d) indicates
that the two channel images overlapped very well, confirming
that Lyso-V can specifically localize in the lysosomes of living
cells. The changes in the intensity profile of linear regions of
2904
dx.doi.org/10.1021/ja311688g | J. Am. Chem. Soc. 2013, 135, 2903−2906

Journal of the American Chemical Society
Communication
the lysosomes and the negligible background fluorescence
outside them.
In combination with FLIM, Lyso-V is the first reliable tool
for determining the local viscosity within lysosomes. Figure 5b
Figure 5. FLIM investigation of an MCF-7 cell stained with 5 μM
Lyso-V. (a) Bright-field image. (b) Fluorescence lifetime imaging
using fluorescence detection at 535 15 nm after pulsed excitation at
405 nm. (c) Histogram of lifetimes.
Figure 3. Colocalization imaging of MCF-7 cells stained with 5.0 μM
Lyso-V and 5.0 μM NR for 3 min at 37 °C. (a) Bright-field image. (b)
Confocal image from Lyso-V on channel 1 (λex = 488 nm, λem = 515−
545 nm). (c) Confocal image from NR on channel 2 (λex = 559 nm,
λem = 585−610 nm). (d) Merged image of channels 1 and 2. (e)
Intensity profile of ROIs across MCF-7 cells. (f) Correlation plot of
Lyso-V and NR intensities.
shows a FLIM image of a single normal MCF-7 cell, in which
the fluorescence lifetimes in the lysosomes have been mapped
with considerable spatial resolution. The FLIM image also
shows a narrow lifetime distribution range between 1.75 and
2.35 ns (Figure 5c). In the lifetime histogram, the maximum
probability is located at 2.05 ns. According to our calibration
graph in Figure 2c, the lifetimes of Lyso-V in lysosomes
correspond to the viscosity range 50−90 cP, and the average
viscosity is ∼65 cP. To our knowledge, these fundamental data
of importance for cellular biology have not been available
previously.
Lyso-V has practical applicability in real-time monitoring of
dynamical changes in lysosomal viscosity under stimulation by
different mechanisms. To confirm this, we adopted two kinds of
medicines commonly used in the clinic. First was dexametha-
sone, an anti-inflammatory and immunosuppressant drug that
works as a stabilizer of lysosomal membrane15a and an inhibitor
of lysosomal enzymatic release.15b During a 20 min period of
stimulation using a low dose (5 μM) of dexamethasone,
continuous FLIM images revealed gradually extended lifetimes
within the lysosomes (Figure 6). According to the lifetime
histograms listed in the SI, the average lifetime in the total cell
increased from 1.99 to 2.24 ns, which means that the average
lysosomal viscosity increased from 67 to 85 cP. We also
checked the lifetimes of two distinct regions of lysosomes
(marked as 1 and 2 in Figure 6a) and found a similar tendency
of the lifetime increases (Figure 6h). Obviously, dexamethasone
interest (ROIs) (Lyso-V and NR costaining) tended toward
synchronization (Figure 3e). From the intensity correlation
plots (Figure 3f), a high Pearson’s coefficient and overlap
coefficient of 0.915 and 1.616, respectively, were obtained. The
above colocalization investigation confirmed that Lyso-V can
stain lysosomes specifically.
The capability of Lyso-V to trace lysosomes with high spatial
and temporal resolution was also demonstrated (Figure 4).
Figure 4. (a−d) Confocal images of an MCF-7 cell stained with 5 μM
Lyso-V and stimulated using 3 μM chloroquine. Different pseudo-
colors are used to illustrate the fluorescence images at different
stimulation times of 0, 1, 2, and 5 min. (e−g) Merges of images at two
different times: (e) 0 and 1 min, (f) 1 and 2 min, and (g) 2 and 5 min.
(h) Bright-field image.
MCF-7 cells were stimulated by a low dose (3 μM) of
chloroquine, a typical lysosomal toxicant, in order to drive
lysosomal migration without inducing any other apparent
disturbance in the cells. A time series of confocal microscopy
images with the aid of Lyso-V was recorded in a short period of
5 min. These pictures at different times (Figure 4a−d) revealed
that the lysosomes largely kept their original shapes but their
locations changed little by little. Although these short-term
shifts were not very large, they could be unambiguously traced
in the merges of images at any two continuous times (Figure
4e−g). Such a good image quality of Lyso-V for tracing
lysosomes should be ascribed to the bright fluorescence within
Figure 6. Fluorescence lifetime imaging of a MCF-7 cell stained with 5
μM Lyso-V and stimulated for different times (0−20 minutes) using 5
μM dexamethasone, with fluorescence detection at 535 15 nm. (a)
Bright-field image. (b−g) Fluorescence lifetime imaging of the cell
stimulated for different times usinig 5 μM dexamethasone. (h) Plot of
fluorescence lifetimes of Lyso-V stimulated for different times using 5
μM dexamethasone.
2905
dx.doi.org/10.1021/ja311688g | J. Am. Chem. Soc. 2013, 135, 2903−2906

Journal of the American Chemical Society
Communication
(3) (a) Haidekker, M. A.; Theodorakis, E. A. Org. Biomol. Chem.
2007, 5, 1669. (b) Kaliviotis, E.; Yianneskis, M. Proc. Inst. Mech. Eng.,
Part H 2007, 221, 887.
(4) Kuimova, M. K.; Botchway, S. W.; Parker, A. W.; Balaz, M.;
Collins, H. A.; Anderson, H. L.; Suhling, K.; Ogilby, P. R. Nat. Chem
2009, 1, 69.
(5) (a) Dell’Angelica, E. C.; Mullins, C.; Caplan, S.; Bonifacino, J. S.
FASEB J. 2000, 14, 1265. (b) Burleigh, M. C.; Barrett, A. J.; Lazarus,
G. S. Biochem. J. 1974, 137, 387.
(6) (a) Futerman, A. H.; van Meer, G. Nat. Rev. Mol. Cell Biol. 2004,
5, 554. (b) Vellodi, A. Br. J. Haematol. 2005, 128, 413. (c) Neufeld, E.
F. Annu. Rev. Biochem. 1991, 60, 257. (d) Simonsa, K.; Gruenberg, J.
Trends Cell Biol. 2000, 10, 459.
(7) (a) Harkness, J. Biorheology 1971, 8, 171. (b) Huizing, M.; Helip-
Wooley, A.; Westbroek, W.; Gunay-Aygun, M.; Gahl, W. A. Annu. Rev.
Genomics Hum. Genet. 2008, 9, 359.
(8) (a) Haidekker, M. A.; Brady, T. P.; Lichlyter, D.; Theodorakis, E.
A. J. Am. Chem. Soc. 2006, 128, 398. (b) Haidekker, M. A.; Ling, T.;
Anglo, M.; Stevens, H. Y.; Frangos, J. A.; Theodorakis, E. A. Chem.
Biol. 2001, 8, 123.
(9) (a) Kuimova, M. K.; Yahioglu, G.; Levitt, J. A.; Suhling, K. J. Am.
Chem. Soc. 2008, 130, 6672. (b) Levitt, J. A.; Kuimova, M. K.;
Yahioglu, G.; Chung, P. H.; Suhling, K.; Phillips, D. J. Phys. Chem. C
2009, 113, 11634.
(10) Peng, X. J.; Yang, Z. G.; Wang, J. Y.; Fan, J. L.; He, Y. X.; Song,
F. L.; Wang, B. S.; Sun, S. G.; Qu, J. L.; Qi, J.; Yan, M. J. Am. Chem.
Soc. 2011, 133, 6626.
stimulation results in an increase in lysosomal viscosity. In view
of the fact that these cells’ lysosomal viscosity before
stimulation was only ∼67 cP, the increase in magnitude of
∼20 cP is remarkable. The other stimulating agent was
chloroquine, which is a typical toxicant for lysosomes because
its stronger basicity can raise the lysosomal pH.16 However,
even at chloroquine concentrations as high as 20 μM, the FLIM
images revealed that the fluorescence lifetimes within the
lysosomes changed very slightly (see the SI). The average
lifetime after chloroquine stimulation for 20 min was ∼2.06 ns,
which is almost the same as the original lifetime (2.05 ns).
Thus, under our stimulation conditions, chloroquine stim-
ulation did not greatly influence the lysosomal viscosity, in
contrast to dexamethasone stimulation. The above experiments
demonstrate that application of Lyso-V in FLIM to monitor
lysosmal viscosity changes will be a valuable method in
lysosome-related pharmacology and toxicology studies.
In conclusion, we have developed Lyso-V, the first
fluorescent probe for lysosomal viscosity, in a straightforward
manner by attaching a morpholine moiety to a BODIPY rotor.
Because of its lysosome-activated fluorescence characteristics,
Lyso-V is an ideal lysosomal tracer with high spatial and
temporal resolution under laser confocal microscopy. More
importantly, Lyso-V can be practically applied in real-time
quantification of lysosomal viscosity changes in live cells
through the FLIM technique. For the first time, lysosomal
viscosity data for MCF-7 cells have been obtained and the
different influences of two typical lysosome-stimulating drugs
on the lysosomal viscosity have been discovered.
(11) Yu, H. B.; Xiao, Y.; Jin, L. J. J. Am. Chem. Soc. 2012, 134, 17486.
(12) de Silva, A. P.; Gunaratne, H. Q. M.; Gunnlaugsson, T.; Huxley,
A. J. M.; McCoy, C. P.; Fademacher, J. T.; Rice, T. E. Chem. Rev. 1997,
97, 1515.
(13) Doose, S.; Neuweiler, H.; Sauer, M. ChemPhysChem 2009, 10,
1389.
(14) Berezin, M. Y.; Achilefu, S. Chem. Rev. 2010, 110, 2641.
(15) (a) Hinz, B.; Hirschelmann, R. Pharm. Res. 2000, 17, 1489.
(b) Ackerman, N. R.; Beebe, J. R. J. Pharmacol. Exp. Ther. 1975, 193,
603.
(16) Gonzalez-Noriega, A.; Grubb, J. H.; Talkad, V.; Sly, W. S. J. Cell
Biol. 1980, 85, 839−852.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis, characterization, and experimental details. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
xiaoyi@dlut.edu.cn
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by National Natural Science
Foundation of China (21174022 and 21173217), the National
Basic Research Program of China (2013CB733702), and the
Specialized Research Fund for the Doctoral Program of Higher
Education (20110041110009). We also thank Prof. Junle Qu at
Shenzhen University for helpful discussions on FLIM.
REFERENCES
■
(1) (a) Elcock, A. H. Curr. Opin. Struct. Biol. 2010, 20, 196.
(b) Minton, A. P. J. Biol. Chem. 2001, 276, 10577.
(2) (a) Luby, P. K. Int. Rev. Cytol. 2000, 192, 189. (b) Stutts, M. J.;
Canessa, C. M.; Olsen, J. C.; Hamrick, M.; Cohn, J. A.; Rossier, B. C.;
Boucher, R. C. Science 1995, 268, 847. (c) Reinhart, W. H. Biorheology
2001, 38, 203. (d) Uchimuar, I.; Numano, F. Diabetes Front. 1997, 8,
33. (e) Simon, A.; Gariepy, J.; Chironi, G.; Megnien, J. L.; Levenson, J.
J. Hypertension 2002, 20, 159. (f) Moriarty, P. M.; Gibson, C. A.
Cardiovasc. Rev. Rep. 2003, 24, 321. (g) Bosman, G. J. C. G. M.;
Bartholomeus, I. G. P.; Grip, D. W. J. Gerontology 1991, 37, 95.
2906
dx.doi.org/10.1021/ja311688g | J. Am. Chem. Soc. 2013, 135, 2903−2906