Fluorescent probe for early mitochondrial voltage dynamics

Article abstract of DOI:10.1039/d1cc01944a

Mitochondrial voltage dynamics plays a crucial role in cell healthy and disease. Here, a new fluorescent probe to monitor mitochondrial early voltage variations is described. The slowly permeant probe is retained in mitochondria during measurements to avoid interferences from natural membrane potential by incorporating an hydrolizable ester function. Voltage, local polarity, pH parameters and transmembrane dynamics were found to be deeply correlated opening a approach in mitochondrial sensing.

Full text of DOI:10.1039/d1cc01944a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Fluorescent probe for early mitochondrial voltage
dynamics†
Cite this: Chem. Commun., 2021,
57, 5526
Cinthia Hernandez-Juarez, Ricardo Flores-Cruz and Arturo Jimenez-Sanchez
*
´
´
´
´
Received 12th April 2021,
Accepted 4th May 2021
DOI: 10.1039/d1cc01944a
rsc.li/chemcomm
Mitochondrial voltage dynamics plays a crucial role in cell healthy commonly, intramolecular electron transfer (which is voltage
and disease. Here, a new fluorescent probe to monitor mitochon- sensitive) is coupled with a proton transfer process.
drial early voltage variations is described. The slowly permeant
We hypothesized that, by designing a probe that features
probe is retained in mitochondria during measurements to avoid a local polarity correlation with DCm could allow a self-
interferences from natural membrane potential by incorporating an referenced monitoring of the mitochondrial voltage variations
hydrolizable ester function. Voltage, local polarity, pH parameters with local polarity through the calibrated fluorescence profile
and transmembrane dynamics were found to be deeply correlated of the probe, where such profile can easily be controlled by a
opening a approach in mitochondrial sensing.
photoinduced electron transfer (PET) process inherent of this
family of molecular polarity probes.6,7
Mitochondrial membrane potential (DCm) is now recognized as
In this work we describe the first molecular probe, called
a universal indicator of mitochondrial function in health and RVolt, to monitor early mitochondrial voltage variations through
disease.1 DCm status is crucial for normal cell growth and a polarity-referenced and non-releasable mechanism, Fig. 1.
differentiation, its disruption however, is directly associated to
all types of mitochondrial-related diseases such as cancer,
diabetes mellitus, cardiovascular and neurodegenerative disor-
ders, to mention some of them.2–4 This membrane potential (or
voltage) is mainly a consequence of the ionic gradients between
the mitochondrial membrane and cytosol,5 and the proper
monitoring of early voltage variations is highly important for
the understanding of the molecular basis that govern cell
processes in mitochondrial disfunction. However, although
several methods have been developed to quantify DCm, most
of them are necessarily of the Nernstian behaviour where
cationic-lipophilic dyes accumulate in the organelle through
the electrochemical gradient (mitochondria has a DCm ca.
À180 mV at pH about 8), then several analytical interferences
arise as a consequence of dye concentration uptake variations
due to mitochondrial depolarization (less negative potential)
and hyperpolarization (more negative potential). Another key
requirement for DCm monitoring is that potential variations
should be calibrated vs. a fluorescence process directly related
to the photophysical mechanism giving the voltage response,
´
´
´
Instituto de Quımica, Universidad Nacional Autonoma de Mexico, Ciudad
Fig. 1 (A) Chemical structure of the RVolt and Control probes. Co-
localization imaging in live HeLa cells under the using MitoLiteTM Blue
(blue panels, lexc = 344 nm, lem = 469 nm) indicating that (B) RVolt is
mitochondria-specific with Pearson’s coefficient of 0.94 and (C) Control is
a non-permeable exogenous probe. Scale bars represent 20 mm.
´
´
Universitaria, Circuito Exterior s/n, De. Coyoacan 04510, Ciudad de Mexico,
Mexico. E-mail: arturo.jimenez@iquimica.unam.mx
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d1cc01944a
5526 | Chem. Commun., 2021, 57, 5526–5529
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
simultaneously directs Control mostly to the plasma membrane.
Also, while RVolt has a partition coefficient log P = 1.02 (and
positively charged), Control has a log P = 0.31 (and neutral, pKa =
6.78 for the carboxylic deprotonation equilibrium, see ESI†) which
is still a bit lipophilic due to the large cinnamic fragment tether to
the rhodamine-type structure.
To corroborate the PET process, present in the RVolt,
a robust photophysical analysis was carried out by means of
UV-Vis and fluorescence spectroscopies as well as computa-
tional studies.
The computational analysis through a three-model scheme
of electron density difference,11 natural transition orbital (NTO)
hole-electron distributions12 and charge transfer extent index
upon photoexcitation11 for the RVolt indicated efficient charge
transfer features, Fig. S2 (ESI†). Charge-transfer (CT) indexes
upon photoexcitation and NTO eigenvalues are summarized in
Fig. S2 and Table S1 (ESI†). Inspection of NTO through the
HONTO–LUNTO pairs corroborated a dipolar charge transfer
redistribution for both probes. Such solvent polarity and pro-
ticity dependence on the fluorescence properties is traduced in
a dual-emission band pattern that allowed us to calibrate the
probe by using the l-ratiometric principle, Fig. 2.
Scheme 1 Sensing mechanism of RVolt to monitor early mitochondrial
membrane voltage. The probe is internalized and retained in mitochondria
through esterase hydrolysis losing its Nernstian features. Then, the PET
process is not active in the hyperpolarized membrane which gives
more fluorescence, depolarization activates PET and quench the NIR
fluorescence. This can be achieved by using carbonyl cyanide
3-chlorophenylhydrazone (CCCP) type uncouplers.
As organic solvents can exert a strong influence on the probe
photophysics by different empirical parameters (i.e., solvent
polarity, polarizability, acidity, basicity and viscosity) at the
same time, proper calibration should be conducted in a more
controlled media, trying to isolate the effect of dielectric con-
stant as much as possible. Then, as dioxane and water have
similar dispersion properties but highly different dipolarities,
we implemented a dioxane:H2O fluorescence titration, Fig. 2,
although the solvatochromic analysis in different solvents for
RVolt is also presented in Table S2 (ESI†). In addition to the
dioxane:H2O continuous variation, we implemented a carbonyl
cyanide 3-chlorophenylhydrazone (CCCP) titration in 5 mM
HTAB micellar medium in order to calibrate both effects. For
both, dioxane:H2O and CCCP, protonation plays the strongest
effect where CCCP promotes a ratiometric change of the Vis
and NIR-emission bands corroborating the PET process inhibi-
tion, overall less fluorescence emission detected at the NIR
band due to the protonation of the dimethylamino(cinnamic)
This implies that after loading cells with the Nernstian fluoro-
phore RVolt having an acetyl function, mitochondrial membrane
is permeabilized. Once the probe is accumulated in the mitochon-
drial matrix, esterase enzymes could hydrolyze the acetyl ester
function,8 disbalancing the formal charge of the probe and thus
trapping RVolt between the mitochondrial matrix and the inner
membrane. Consequently, the hydrolyzed probe is no longer
permeable to the membrane, similar to the slow-equilibrating
mitochondrial marker Rhodamine 123,9 Scheme 1. In fact, RVolt
is similar to Rhodamine 123 having slowly permeant features,10
and larger periods of mitochondrial retention, thus allowing
almost 1 hour before leaving the mitochondrial outer membrane.
Importantly, as RVolt is highly dependent on the local polarity, its
fluorescent profile calibration towards this physicochemical
observable allowed us to have an internal reference to monitor
mitochondrial voltage in live cells under conditions promoting
membrane potential variations, swelling, permeability transition
pore (mPTP) opening and under pH and redox variation controls.
RVolt was synthetized and characterized as described in the
ESI.† Then, photophysical properties and high-resolution con-
focal microscopy studies in HeLa and SK-Lu-1 cells were carried
out. The chemical structures and subcellular localization of
RVolt and an RVolt-derived control, named Control, are pre-
sented in Fig. 1A. Specific mitochondrial localization was
corroborated by co-localization analysis for RVolt using the
commercial blue-emitter MitoLitet marker, Fig. 1B. A highly
similar co-localization pattern was also found in HK-Lu-1 cells,
Fig. S1 (ESI†). Yet normal incubation (exogenous addition) of
Control molecule to the cells does not lead to a high cell
permeabilization and mitochondrial staining. This can be
explained by the presence of a free carboxylate function that
neutralizes the oxonium-positive charge of the fluorophore and
Fig. 2 CCCP and polarity model parameters. (A) Fluorescence spectra
(lexc = 400 nm) of 40 mM RVolt at variable CCCP concentration. Inset
shows fluorescence spectra of RVolt at different dioxane:H2O v/v mix-
tures. (B) Ratiometric calibration plot for fluorescence intensity ratio
r = 510/715 nm (straight line, middle) at the corresponding CCCP [nM]
(filled circles, left) and correlation with % dioxane v/v (hollow circles, right).
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 5526–5529 | 5527

View Article Online
ChemComm
Communication
moiety. The dioxane:H2O variation introduces a ratiometric activated upon CCCP protonophore additions, thus resulting in
pattern although a stronger activation at the Vis emission band a decrease in fluorescence intensity. To assess the membrane
was observed. Then, the ratiometric calibration plot for r = 510/ voltage variation monitoring such fluorescence scenario was
715 nm at the corresponding CCCP [nM] and corresponding % calibrated and correlated with the fluorescence imaging in cells
dioxane v/v variation is presented in Fig. 2B. Such calibration by tracking membrane depolarization and hyperpolarization.
provides an excellent tool to correlate the proticity effect com- Another feature of the RVolt is that the obtained fluorescence
ing from the CCCP uncoupler and other protonation effects intensities can be directly associated with the intrinsic dipolar
that can occur in the subcellular environment, i.e. osmotically nature of the probe. For this reason, the l-ratiometric profile of
driven water into the membranes. On the other side, high- RVolt vs. CCCP has a strong linear dynamic range and sensi-
resolution confocal imaging using an Airyscan detector for tivity, Fig. 2B. It is worth mentioning that PET process in this
HeLa and SK-Lu-1 cells stained with the RVolt revealed a clean type of donor–acceptor fluorophores is always inherently asso-
mitochondrial framework localization. Such filamentous-like ciated with a charge transfer state, in this sense is crucial for
morphology was visualized in both, the green (lexc = 488 nm) any analytical calibration to consider the extent to which local
and red (lexc = 647 nm) confocal channel setup during the polarity influence the observed fluorescence profile.
experiments. Interestingly, confocal imaging studies using
We pursued an interference-free scenario of the RVolt under
CCCP uncoupling agent to depolarize mitochondria demon- a controlled mitochondrial pH using a previously reported
strated that once RVolt is internalized in the organelle, the protocol by our group,14 the effect of the mitochondrial
mitochondrial localization persisted, although with variable membrane potential variations was analyzed using CCCP and
fluorescence intensity from low-dose experiments to high oligomycin A (OA) which increases DCm by inhibition of ATP
concentration of CCCP (150 nM to 500 nM), Fig. 3A. In order synthase, leading to a decrease in the reduced respiratory chain
to demonstrate that RVolt follows a non-Nernstian mitochon- intermediates. Thus, we turned to exploring the influence of pH
drial localization mechanism, we conducted three key experi- on mitochondrial membrane voltage reversibility by using a
ments, to know, (1) the fluorescence titration of RVolt at 100 nM nigericin stimuli before the depolarization (with CCCP)
variable CCCP concentration to corroborate the observed and hyperpolarization (with OA) treatments.15 The mitochon-
fluorescence intensity variation, (2) cell imaging reversibility drial pH gradient (DpHm) was substantially abolished since
study using CCCP–oligomycin A stimuli, (3) a mitochondrial nigericin promotes a K+/H+ membrane equilibration making
permeability transition pore (mPTP) opening study to confirm [K+]in = [K+]out, consequently the proton concentration [H+]in
=
that the RVolt is trapped between the mitochondrial matrix and [H+]out and DpHm = 0. Live-cell experiments with the RVolt
the inner membrane. During the experiments, we used the probe were always corrected by this control parameter as a
commercial TMRMs mitochondrial localizer as control, starting point. To note, oligomycin increases mitochondrial pH
Fig. 3B, although this dye exhibits self-quenching at high (less [H+]m), while CCCP decreases mitochondrial pH (more
concentrations, thus interfering the interpretation of fluores- [H+]m),16 Fig. 4. As previous cautionary case studies,17 it is
cence intensity changes.
important to consider that positive charges in the membrane
As mentioned, PET process is the photophysical mechanism voltage dynamics are not only protonic charges as usually
that allows the probe to report membrane voltage variations interpreted, but also K+ and Ca2+ charges dumping from
without leaving the mitochondria. As protonophores, CCCP or mitochondria into the cytosol, for such reason even before
FCCP agents depolarize the mitochondrial inner membrane
thus increasing the PET process13 and changing the fluores-
cence intensity profile of RVolt, Fig. 2 and 3A. As shown, PET is
Fig. 4 Induced mitochondrial voltage changes in HeLa cells. Relative
fluorescence intensity (Fi À F0 Æ SEM in r.a.u.) vs. time (min) for 1 mM
RVolt (green and red lines) and TMRMs mitochondrial localizer
Fig. 3 Time course monitoring of CCCP depolarization changes. (A) Cells
are stained with 10 mM RVolt and 150 nM CCCP were added. Then, images
were recorded after each 5 min with corresponding CCCP additions until
reaching 500 nM (lexc = 488 and 647 nm). (B) TMRM mitochondrial
localizer control experiment (lexc = 605 nm). Scale bars = 20 mm.
control (orange line) upon in situ additions of 100 nM nigericin, 150 nM
CCCP (after 5 min) and 5 mg mLÀ1 oligomycin A at 20 min. SLM-
spectrophotometer setup (lexc/lem): green (488/510 nm), red (700/
715 nm) and orange (605/615 nm). Point recordings were taken by 1 min
time-lapse.
5528 | Chem. Commun., 2021, 57, 5526–5529
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
red confocal channels) and Fig. 5 indicates that Control probe
and the in situ esterase hydrolyzed RVolt probe released by
mPTP opening have highly similar imaging patterns, that is, a
cytoplasmic localization for both thus suggesting that the RVolt
can be released by the activation of the mPTP opening process
and not by a Nernstian mechanism.
The RVolt is a powerful new tool for the monitoring of
mitochondrial membrane voltage variations. With a simple
electronic dipolar fluorophore as a local polarity reporter, the
PET process was successfully correlated with voltage variations
and local polarity and under mitochondrial pH control to
isolate the fluorescent profiles from factors other than voltage
variations. The mPTP opening study is critical to corroborate
the non-Nernstian mechanism of voltage probes. The present
strategy may find utility in the design and development of new
fluorescent probes with slow permeability for multiplexing
assays correlating membrane voltage and physicochemical
parameters such as local viscosity, polarity and redox status
at the subcellular level.
Fig. 5 Effect of mPTP opening detected in the deep-red confocal chan-
nel (lexc = 647 nm) for HeLa cells before (first panel-left) and after 5 min
treatment with 400 mM CaCl2 (panels 2–6). The images were taken each
2 minutes and were manually focused, excitation light was fully shielded
between recordings to prevent artefacts and photobleaching. Scale bars =
20 mm.
CCCP stimuli, nigericin does not completely inhibited mito-
chondrial depolarization (the slight fluorescence increments
after nigericin treatment in Fig. 4). Interestingly, this subtle
effect was successfully monitored by the RVolt, fluorescence
profiles in the green and red lines of Fig. 4. Then, the elicited
responses upon treatment with CCCP and OA were corrobo-
rated as expected, red and green profiles in Fig. 4. As can be
seen, CCCP increases the rate of PET process in the RVolt
causing a fluorescence decrease in the red confocal channel
and a slight increment (almost constant) of the fluorescence
intensity in the green confocal channel. The OA treatment that
slows down such electron transfer resulted in the RVolt fluores-
cence activation inside the organelle. Since the mitochondrial
membrane potential is only a part of the transmembrane
potential energy coming from the proton gradient (DmH+) on
the inner mitochondrial membrane,18 regulation of the proton
gradient is desirable. In fact, experiments without nigericin
treatment produced notorious mitochondrial damage and
noisy fluorescence variations just after CCCP treatment.
Conflicts of interest
There are no conflicts to declare.
Notes and references
1 A. A. Bazhin, R. Sinisi, U. Marchi, A. Hermant, N. Sambiagio,
T. Maric, G. Budin and E. A. Goun, Nat. Chem. Biol., 2020, 16, 1385.
2 Y. Wang, E. Xu, P. R. Musich and F. Lin, CNS Neurosci. Ther., 2019,
25, 816.
3 V. A. Popkov, E. Y. Plotnikov, D. N. Silachev, I. B. Pevzner,
S. S. Jankauskas, V. A. Babenko, S. D. Zorov, A. V. Balakireva,
M. Juhaszova, S. J. Sollott and D. B. Zorov, Anal. Biochem., 2018,
552, 50.
4 M. Momcilovic, et al., Nature, 2019, 575, 380. See ESI† for the full
name list.
5 M. Grabe and G. Oster, J. Gen. Physiol., 2001, 117, 329.
6 X. Tian, L. C. Murfin, L. Wu, S. E. Lewis and T. D. James, Chem. Sci.,
2021, 12, 3406.
7 D. C. Magri, Coord. Chem. Rev., 2021, 426, 213598.
8 G. A. Rutter, P. Burnett, R. Rizzuto, M. Brini, M. Murgia, T. Pozzan,
Finally, we envisioned that a promising strategy to release
the hydrolyzed RVolt probe from the mitochondrial interior,
thus corroborating that the probe is trapped inside the orga-
nelle by a non-Nernstian mechanism should involve the open-
ing of the mitochondrial permeability transition pore (mPTP)
´
J. M. Tavare and R. M. Denton, Proc. Natl. Acad. Sci. U. S. A., 1996,
93, 5489.
9 J. J. Lemasters and V. K. Ramshesh, Methods Cell Biol., 2007, 80, 283.
and not through a diffusion driven by CCCP depolarization. 10 H. Rottenberg and S. Wu, Biochim. Biophys. Acta, 1998, 1404, 393.
´
11 M. Savarese, C. A. Guido, E. Bremond, I. Ciofini and C. Adamo,
Firstly, we conducted a mass spectrometry RVolt titration with
esterase (carboxyl ester hydrolase) to assess the hydrolysis
J. Phys. Chem. A, 2017, 121, 7543.
12 R. L. Martin, J. Chem. Phys., 2003, 118, 4775.
reaction occurring from RVolt to Control probe. Fig. S3 (ESI†) 13 P. E. Z. Klier, J. G. Martin and E. W. Miller, J. Am. Chem. Soc., 2021,
143, 4095.
shows that Control probe is in fact formed upon esterase
additions, corroborating the effectiveness of this hydrolizable
´
´
14 R. Flores-Cruz and A. Jimenez-Sanchez, Chem. Commun., 2018,
54, 13997.
function, similar to the Calcein AM protocol (MitoProbet 15 S. W. Perry, J. P. Norman, J. Barbieri, E. B. Brown and H. A. Gelbard,
Transition Pore Assay Kit).19 Then, experiments with the mPTP
opening resulted in the release of the hydrolyzed RVolt probe,
Biotechniques, 2011, 50, 98.
16 J. P. Norman, S. W. Perry, K. A. Kasischke, D. J. Volsky and
H. A. Gelbard, J. Immunol., 2007, 178, 869.
Fig. 5. mPTP opening is a non-selective mechanism by which 17 J. P. Norman, S. W. Perry, H. M. Reynolds and M. Kiebala, et al., PLoS
One, 2008, 3(11), e3731.
18 D. Zorova, V. A. Popkov, E. Y. Plotnikov and D. N. Silachev, et al.,
small molecules pass out the mitochondrial membrane, for this
reason we envisioned that a control experiment to demonstrate
Anal. Biochem., 2018, 552, 50.
that the hydrolyzed version of the fluorophore is the one being 19 MitoProbet Transition Pore Assay Kit (M34153). Molecular probes,
2007. For a representative reference please see: A. J. Kowaltowski,
S. S. Smaili, J. T. Russell and G. Fiskum, Am. J. Physiol.: Cell Physiol.,
released, would be critical. Treatment with 400 mM CaCl2 was
used to selectively induce mPTP opening without causing
2000, 279, C852.
organelle swelling.20 Comparison between Fig. 1C (green and 20 M. Panel, B. Ghaleh and D. Morin, Sci. Rep., 2017, 7, 4283.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 5526–5529 | 5529