A dual-fluorophore sensor approach for ratiometric fluorescence imaging of potassium in living cells

Article abstract of DOI:10.1039/d0sc03844j

Potassium is the most abundant intracellular metal in the body, playing vital roles in regulating intracellular fluid volume, nutrient transport, and cell-to-cell communication through nerve and muscle contraction. On the other hand, aberrant alterations in K⁺homeostasis contribute to a diverse array of diseases spanning cardiovascular and neurological disorders to diabetes to kidney disease to cancer. There is an unmet need for studies of K⁺physiology and pathology owing to the large differences in intracellularversusextracellular K⁺concentrations ([K⁺]_intra= 150 mM, [K⁺]_extra= 3-5 mM). With a relative dearth of methods to reliably measure dynamic changes in intracellular K⁺in biological specimens that meet the dual challenges of low affinity and high selectivity for K⁺, particularly over Na⁺, currently available fluorescent K⁺sensors are largely optimized with high-affinity receptors that are more amenable for extracellular K⁺detection. We report the design, synthesis, and biological evaluation of Ratiometric Potassium Sensor 1 (RPS-1), a dual-fluorophore sensor that enables ratiometric fluorescence imaging of intracellular potassium in living systems.RPS-1links a potassium-responsive fluorescent sensor fragment (PS525) with a low-affinity, high-selectivity crown ether receptor for K⁺to a potassium-insensitive reference fluorophore (Coumarin 343) as an internal calibration standard through ester bonds. Upon intracellular delivery, esterase-directed cleavage splits these two dyes into separate fragments to enable ratiometric detection of K⁺.RPS-1responds to K⁺in aqueous buffer with high selectivity over competing metal ions and is sensitive to potassium ions at steady-state intracellular levels and can respond to decreases or increases from that basal set point. Moreover,RPS-1was applied for comparative screening of K⁺pools across a panel of different cancer cell lines, revealing elevations in basal intracellular K⁺in metastatic breast cancer cell linesvs.normal breast cells. This work provides a unique chemical tool for the study of intracellular potassium dynamics and a starting point for the design of other ratiometric fluorescent sensors based on two-fluorophore approaches that do not rely on FRET or related energy transfer designs.

Full text of DOI:10.1039/d0sc03844j

Chemical
Science
View Article Online
View Journal
EDGE ARTICLE
A dual-fluorophore sensor approach for ratiometric
fluorescence imaging of potassium in living cells†
Cite this: DOI: 10.1039/d0sc03844j
a
abc
Zeming Wang,a Tyler C. Detomasi
and Christopher J. Chang
*
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
Potassium is the most abundant intracellular metal in the body, playing vital roles in regulating intracellular
fluid volume, nutrient transport, and cell-to-cell communication through nerve and muscle contraction. On
the other hand, aberrant alterations in K+ homeostasis contribute to a diverse array of diseases spanning
cardiovascular and neurological disorders to diabetes to kidney disease to cancer. There is an unmet
need for studies of K+ physiology and pathology owing to the large differences in intracellular versus
extracellular K+ concentrations ([K+]intra ¼ 150 mM, [K+]extra ¼ 3–5 mM). With a relative dearth of
methods to reliably measure dynamic changes in intracellular K+ in biological specimens that meet the
dual challenges of low affinity and high selectivity for K+, particularly over Na+, currently available
fluorescent K+ sensors are largely optimized with high-affinity receptors that are more amenable for
extracellular K+ detection. We report the design, synthesis, and biological evaluation of Ratiometric
Potassium Sensor 1 (RPS-1), a dual-fluorophore sensor that enables ratiometric fluorescence imaging of
intracellular potassium in living systems. RPS-1 links a potassium-responsive fluorescent sensor fragment
(PS525) with a low-affinity, high-selectivity crown ether receptor for K+ to a potassium-insensitive
reference fluorophore (Coumarin 343) as an internal calibration standard through ester bonds. Upon
intracellular delivery, esterase-directed cleavage splits these two dyes into separate fragments to enable
ratiometric detection of K+. RPS-1 responds to K+ in aqueous buffer with high selectivity over competing
metal ions and is sensitive to potassium ions at steady-state intracellular levels and can respond to
decreases or increases from that basal set point. Moreover, RPS-1 was applied for comparative screening
of K+ pools across a panel of different cancer cell lines, revealing elevations in basal intracellular K+ in
metastatic breast cancer cell lines vs. normal breast cells. This work provides a unique chemical tool for
the study of intracellular potassium dynamics and a starting point for the design of other ratiometric
fluorescent sensors based on two-fluorophore approaches that do not rely on FRET or related energy
transfer designs.
Received 14th July 2020
Accepted 3rd December 2020
DOI: 10.1039/d0sc03844j
rsc.li/chemical-science
malfunction of K+ channels or pumps, can contribute to serious
diseases that range from neurodegeneration and epilepsy to
Introduction
All cells in all living organisms require potassium,1,2 where the
abundant K+ cation regulates cellular osmolality, nutrient
status, and membrane potential,3–5 contributing to physiolog-
ical processes including neuron ring, muscle contraction,6,7
and hormone regulation.8–10 Cellular potassium homeostasis is
maintained by an intricate network of membrane K+ channels
and pumps, giving physiological K+ concentrations ranging
from approximately 3–5 mM extracellularly to 150 mM intra-
cellularly.11 Aberrant elevations or depletions of intracellular K+
from these tightly regulated set points, caused largely by
kidney failure to diabetes to arrhythmia and heart disorders to
cancer.12,13 In one example, a growing body of literature
connects K+ channels to cancer through cell motility, which is
critical to tumor growth and metastasis.12–21 However, most of
these studies have focused on correlations of aberrant expres-
sion of K+ channels in various types of cancer and suggest
a relationship between altered intracellular K+ level and cell
apoptosis,16,22,23 with limited information on direct measure-
ments of K+ pools themselves.
These foregoing considerations motivate the development of
chemical tools that can enable detection of K+ pools in living
biological specimens with higher spatial and temporal resolu-
tion compared to conventional methods of detection such as K+-
selective microelectrodes and patch clamp techniques.24,25
Notably, recent years have also seen advances in the develop-
ment of genetically encoded potassium probes and nano-
sensors.26–31 In this context, small-molecule uorescent sensors
aDepartment of Chemistry, University of California, Berkeley, CA 94720, USA. E-mail:
chrischang@berkeley.edu
bDepartment of Molecular and Cell Biology, University of California, Berkeley, CA
94720, USA
cHelen Wills Neuroscience Institute, University of California, Berkeley, CA 94720, USA
† Electronic supplementary information (ESI) available. See DOI:
10.1039/d0sc03844j
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
offer an attractive approach to meet this goal, as they are valued detection of K+ in the cytosol without saturation, limiting their
as non-invasive, high signal-to-noise reagents that can accu- utility for studies of intracellular K+ dynamics (Fig. 1).
rately map a variety of biological analytes, particularly metal
In addition, currently available small-molecule uorescent
ions.32–37 Despite notable advances in the development of small- K+ sensors are limited to intensity-based responses. In the best
molecule uorescent sensors for potassium,38–45 imaging intra- case scenario, such turn-on uorescence potassium probes can
cellular K+ in biological samples with high selectivity over Na+ detect dynamic K+ uxes within the same cell population before
remains a signicant and insufficiently solved challenge, owing and aer a stimulation event, but comparing relative K+ levels
in large part to high resting K+ concentrations within the cell across different biological samples can be complicated. Proper
(150 mM), which requires a low-affinity molecular recognition comparisons can be hindered by variations in: intracellular
element that still achieves high K+ selectivity. Pioneering work delivery, uptake, and retention of probes among cell types;
by Tsien and Minta on the rst potassium probe, potassium sample thickness and other heterogeneities, and both excita-
benzofuran isophthalic acid (PBFI), offered an important early tion and emission light intensity and power dependency. These
solution to this problem, but PBFI is limited by slow cellular issues can be mitigated by ratiometric uorescence imaging,
uptake kinetics due to its high molecular weight and need for which enables calibration by an internal standard.54 Along these
UV excitation, which can trigger background autouorescence lines, we have recently reported activity-based sensing probes
and sample photodamage.46 Likewise, He and colleagues for iron55 (FRET Iron Probe-1, FIP-1) and copper56 (FRET Copper
developed a high-affinity uorescent K+ sensor based on a tri- Probe-1, FCP-1), which feature ratiometric uorescence detec-
aza-cryptand potassium receptor,38 and Verkman and other tion using a uorescence resonance energy transfer (FRET)
research teams have successfully developed several useful strategy.
sensors for monitoring extracellular K+, including stimulated K+
Here we report the design, synthesis, and biological eval-
release, based on this ionophore.47–50 Specically, the Verkman uation of a rst-generation ratiometric uorescent sensor for
laboratory has demonstrated visualizing K+ waves in brain K+ detection, termed Ratiometric Potassium Sensor-1 (RPS-1).
cortex using a K+ indicator.51 K+ sensors have also been devel- RPS-1 is composed of
a dual-uorophore system that
oped using native and engineered nucleic acid G-quadruplex combines a K+-responsive rhodol moiety (Potassium Sensor
binding motifs.39,40,52,53 As such, the vast majority of available 525, PS525) and a standard uorophore (Coumarin 343) for
uorescent K+ sensors for biological use achieve selectivity of K+ more suited sensing to intracellular [K+] levels in the 150 mM
over Na+ by high-affinity means. However, the tight-binding range with internal self-calibration. Unlike previously reported
nature of these probes does not allow for a dynamic range of FIP-1 and FCP-1 indicators that rely on FRET, the two uo-
rophores in RPS-1 are spectrally distinct and are not designed
to undergo internal energy transfer. Instead, RPS-1 utilizes
ester linkers that can be cleaved by intracellular esterases to
detach the K+-responsive uorophore from its internal stan-
dard and trap both dyes inside the cell in their carboxylate
forms, an approach inspired by previous work by Woodroofe
and Lippard on ratiometric zinc sensors.57 As a key design
feature to tailor its use for intracellular K+ detection, the PS525
portion of RPS-1 utilizes a lariat-modied crown ether to
achieve a Kd value more in line with intracellular K+ levels
(Fig. 1).42,58,59 Indeed, despite its low affinity for K+, RPS-1
shows sufficient selectivity for K+ over competing biologically
relevant metal ions, including Na+, and is able to monitor
intracellular K+ uxes. Moreover, because ratiometric cali-
bration minimizes interference arising from analyte-
independent phenomena such as sample thickness, hetero-
geneity, variations in light intensity, and dye loading, RPS-1
enables a rapid readout of K+ levels among different cell lines
that can be corroborated by ICP-MS. Interestingly, data from
Fig. 1 Design considerations of extracellular and intracellular potas-
sium pools and potassium-binding receptors for the development of
Ratiometric Potassium Sensor-1 (RPS-1). (a) Comparison of extracel-
these independent measurements reveal that select types of
breast cancer cells possess elevated concentrations of K+
relative to normal breast cell counterparts. In addition to
lular and intracellular potassium pools showing higher resting levels of
K+ inside versus outside the cell. (b) Comparison of high-affinity G-
providing a unique chemical tool for interrogating intracel-
quadruplex and cryptand receptors tailored for extracellular K+ binding lular K+ dynamics, this chemical modication strategy of
versus lariat crown ether receptors that are more amenable for
converting a turn-on probe to a ratiometric probe is broadly
applicable to a variety of uorescent probes for different ana-
lytes provided each uorophore has a non-overlapping emis-
sion spectrum. Taken together, RPS-1 provides key advances in
intracellular K+ binding. The Kd values of both G-quadruplex- and
cryptand-based receptors are well-matched with extracellular [K+]
levels in the 3–5 mM range, whereas the weaker-binding lariat crown
ether ionophores are more suited sensing to intracellular [K+] levels in
the 150 mM range.
terms of visible wavelength excitation and emission to
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
Synthesis and characterization of the K+-responsive turn-on
probe Potassium Sensor 525 (PS525)
minimize background autouorescence and photodamage to
biological samples, a low-affinity K+ receptor that retains high
selectivity over Na+ to better match intracellular [K+] levels in
the 150 mM range, and a ratiometric response to enable
internal self-calibration.
As our design of RPS-1 is comprised of a K+-responsive uo-
rescent moiety linked to an internal dye standard for ratio-
metric calibration, we sought to prepare a new turn-on
uorescent sensor that responds selectively to K+ in the high
intracellular [K+] range of 150 mM. To achieve this goal, we
chose a lariat ether modied aza-crown ether receptor as the K+-
responsive moiety, as it retains high selectivity for K+ over Na+
but exhibits a weaker binding affinity for K+ compared to
cryptand and G-quadruplex moieties (Fig. 1b and 2a). The probe
is designed to operate by a standard photoinduced electron
transfer (PET) mechanism where the electron-rich amine
receptor will quench uorescence by PET, which is alleviated by
binding of the positively charged cation to the receptor.33,60 We
appended this receptor to a rhodol-based structure owing to its
spectral separation from Coumarin 343 and attractive imaging
properties, including good optical brightness and photo-
stability in one- and two-photon modes, as well as a balance
between hydrophilicity and cell-permeability.35,61,62
Results and discussion
Design of Ratiometric Potassium Sensor-1 (RPS-1), a dual-
uorophore probe for ratiometric potassium detection
To develop a ratiometric sensing platform for potassium
tailored for intracellular use, we envisioned a dual-uorophore
system with two dyes, one being potassium-sensitive and one
being potassium-insensitive, which could be delivered in
a 1 : 1 ratio to cells. In principle, the potassium-sensitive
moiety would possess a low-affinity receptor to better match
the high levels of resting intracellular K+ pools in the 150 mM
range and respond by an intensity-based turn-on response,
whereas the potassium-insensitive dye would exhibit a distinct
spectral excitation and emission from the potassium sensor
for internal calibration. We chose Coumarin 343 as the
internal standard uorophore owing to its lack of a uores-
cence response to K+, and we designed and synthesized the
new potassium-responsive dye Potassium Sensor 525 (PS525)
for the other portion of the sensor (Scheme 1a). Coumarin 343
is coupled to PS525 via two ester-based linkers to yield RPS-1
(Scheme 1b). Aer RPS-1 is delivered into the cells, cellular
esterases will cleave the ester bonds to give two separate u-
orophores, PS525 (K+-responsive) and Coumarin 343 (K+-
insensitive), in a 1 : 1 ratio (Scheme 1c). Aer cleavage, both
uorophores are cell-trappable due to their negatively charged
carboxylate groups. As such, the ratio of respective emission
proles for PS525 and Coumarin 343 can determine a readout
of intracellular K+ concentrations.
The synthesis of this K+ responsive moiety, Potassium Sensor
525 (PS525), is shown in Scheme 2. Lariat ether modied aza-
crown ether 1 was synthesized using published procedures42
and then converted to xanthene 2 through a Vilsmeier–Haack
reaction using phosphoryl chloride in N,N-dimethylformamide,
followed by resorcinol condensation with triuoroacetic acid as
the solvent. Rhodol 3 was synthesized by converting the
hydroxyl group on 2 to a triate group, followed by Buchwald–
Hartwig amination to add tert-butoxycarbonyl(Boc)-protected
piperazine group and worked up with a trace amount of tri-
uoroacetic acid. Finally, PS525 was obtained by reacting 3 with
methyl 2-bromoacetate, followed by treatment of lithium
hydroxide in tetrahydrofuran and water mixture.
Scheme 1 Design and action of Ratiometric Potassium Sensor-1 (RPS-1). (a) Molecular structures of K+-insensitive Coumarin 343 and K+-
responsive Potassium Sensor 525 (PS525) portions to generate (b) RPS-1. (c) RPS-1 is cleaved by intracellular esterases back to PS525 and
Coumarin 343, which are both cell-trappable owing to their negatively charged carboxylate motifs. Because PS525 has a turn-on response to K+
whereas Coumarin 343 is non-responsive to K+, ratiometric fluorescence measurements can be made using two-channel imaging and be
correlated with changes in K+.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 2 Turn-on response and potassium selectivity for PS525. (a) Photoinduced electron transfer (PET) sensing and K+/Na+ selectivity mech-
anism for lariat ether sensor PS525. (b) Fluorescence spectra of 2 mM PS525 in HEPES buffer (pH 7.4) containing [K+] at 0, 5, 10, 20, 30, 40, 50, 70,
90, 110, 130, 150, 200 mM. The blue trace is at 0 mM K+ and red trace is at 200 mM K+. Excitation was provided at 525 nm. (c) Fluorescence
intensity changes of PS525 in the presence of competing biological metal ions at the physiological concentrations with (red) or without (blue) K+:
K
+ (150 mM), Na+ (5.0 mM), Mg2+ (2.0 mM), Ca2+ (2.0 mM), Mn2+ (10 mM), Fe2+ (10 mM), Co2+ (10 mM), Ni2+ (10 mM), Cu2+ (10 mM), Zn2+ (10 mM). (d)
Response of the ratio of the fluorescence intensity of PS525 and Coumarin 343 to physiologically relevant K+ levels. 5 mM PS525 and Coumarin
343 (1 : 1) in HEPES buffer (50 mM, pH 7.4) was used for measurement. Peak intensity value of PS525 (552 nm) and Coumarin 343 (490 nm) was
obtained to calculate ratio intensity. Error bars denote SEM, n ¼ 3.
We evaluated the optical properties of PS525 in aqueous 552 nm (Fig. 2b), with a quantum yield increase from 0.03 in the
solution buffered to physiological pH (50 mM HEPES, pH 7.4). absence of K+ to 0.15 in the presence of K+ (Fig. S2†). The Kd
PS525 shows a maximum absorption peak at 525 nm with value of the probe was calculated to be 137 mM based on the K+
calculated absorption coefficients of 3 ¼ 61 100 Mꢀ1 cmꢀ1 and 3 titration response using a Benesi–Hildebrand plot (Fig. S1†),
¼ 65 000 Mꢀ1 cmꢀ1 in the presence of 0 and 200 mM of K+, which is well-matched with intracellular [K+] in the 150 mM
respectively (Fig. S2†). When exposed to K+, PS525 shows a 7- range.
fold turn-on increase in uorescence emission centered at
Scheme 2 Synthesis of PS525.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
We next tested the metal selectivity of PS525 for K+ compared
to a panel of biologically relevant alkali, alkaline earth, and
transition metals (Fig. 2c). Notably, even with its low affinity for
K+, PS525 showed a greater than ve-fold selectivity for K+ over
Na+ at relevant intracellular levels, presaging that it possesses
adequate sensitivity and selectivity for downstream biological
applications. We then tested the sensitivity of the ratio
responsiveness of PS525 and Coumarin 343 to the K+ level
around intracellular K+ concentration (140 mM) (Fig. 2d).
Within the range of 80 mM to 120 mM, linear regression anal-
ysis was performed to calculate limit of detection at 5.8 mM.
Additionally, the pH sensitivity test of both PS525 and
Coumarin 343 demonstrated their usability under physiological
pH (Fig. S3†).
Synthesis of Ratiometric Potassium Sensor-1 (RPS-1) from
PS525 and Coumarin 343 dyes and demonstration of
ratiometric uorescence imaging of K+ levels in living cells
Aer establishing that PS525 is a selective and sensitive uo-
rescent K+-sensing dye unit, we synthesized RPS-1 through
esterication of PS525 with a previously reported coumarin
synthon63 using 2-(3H-[1,2,3]triazolo[4,5-b]pyridine-3-yl)-1,1,3,3-
tetramethyluronium tetrauoroborate (TATU) as a carboxylate
activator (Fig. 3a). We tested the uorescence response and
esterase-catalyzed hydrolysis of RPS-1 (Fig. S4 and S5†). We next
applied RPS-1 for ratiometric uorescence imaging of intracel-
lular K+ pools with confocal microscopy, given the ability of the
PS525 sensing unit to respond selectively to K+ levels at bio-
logically relevant concentrations in aqueous buffer in vitro. We
note that the K+-insensitive Coumarin 343 piece has a quantum
yield of 0.02, which is comparable to the PS525 portion, but is
excited at a separate 458 nm wavelength (Fig. S2†). HeLa cells
were rst incubated with 10 mM RPS-1 before imaging to ensure
probe uptake and esterase cleavage. Dual-color confocal
microscopy images were acquired using 458 nm excitation of
Fig. 3 (a) Synthesis of RPS-1 and ratiometric imaging of intracellular
K+ pools using RPS-1. (b) Time-dependent ratiometric fluorescence
imaging with RPS-1 enables monitoring of depletion of intracellular K+
pools in live HeLa cells treated with 5 mM valinomycin for 1 h. The green
channel shows emission from the PS525 chromophore with 514 nm
the Coumarin 343 unit (blue channel) and 514 nm excitation of excitation, and the blue channel shows emission from the Coumarin
343 chromophore with 458 nm excitation. Fluorescence ratio images
from the two channels were constructed from ImageJ. (c) Average
the PS525 unit (green channel). A ratio image was then obtained
using these two channels (green channel/blue channel, Fig. 3b).
To assess whether RPS-1 can respond to changes in intracellular
K+ changes through its PS525 unit, HeLa cells were then treated
fluorescence intensity ratios from multiple biological replicates plotted
as measured and analysed by ImageJ. Decreases of intracellular K+
levels upon valinomycin stimulation are clearly shown by fluorescence
with 5 mM valinomycin, a compound known to chelate and ratio changes (red trace) relative to vehicle control (blue trace). Error
bars denote SEM for mean green/blue ratios, n ¼ 3. Scale bar ¼ 25 mm.
transport potassium selectively through membranes and
deplete intracellular pools of K+. Confocal images from both
channels were taken at 0, 5, 15, 30, 45, 60 min aer valinomycin
addition and compared with a vehicle control. We observed
a clear decrease in green/blue ratio intensities in the HeLa cells
stimulated by valinomycin compared to the cells in the vehicle
control group (Fig. 3c). Furthermore, we conrmed that cells
treated with valinomycin indeed have decreased levels of
intracellular K+ through inductively coupled plasma mass
spectrometry measurements (ICP-MS; Fig. S5†). These data
suggest that RPS-1 can respond to dynamic changes in intra-
cellular K+ pools and is capable of monitoring decreases in [K+]
from basal levels. Moreover, the dual-uorophore ratio from the
vehicle control remains stable throughout the time course of
the experiment, consistent with the design in which the two dye
components are delivered and trapped upon esterase cleavage.
Finally, cell viability was measured through propidium iodide
staining and was not signicantly different with valinomycin
and/or RPS-1 treatments over the course of the experiment
(Fig. S6†).
We next tested the ability of RPS-1 to detect increases in
intracellular K+ levels upon stimulation. HT29 cells are known
to selectively express signicantly higher levels of the human
`
ether-a-go-go-related gene (HERG) potassium channel relative
to other cancer cell lines such as HeLa and A549.64,65 We treated
HT29 cells with astemizole, a known HERG potassium channel
blocker, to enable expansion of intracellular K+ pools.66 We rst
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 4 Application of RPS-1 to assay astemizole-induced K+ influx
across a variety of cell lines. Data shown for HT29 cancer cells
compared to HeLa and A549 cells under basal conditions and after
treatment with 5 mM asterizole for 1 h. Mean Green/Blue ratio was
obtained for each cell line; error bars denote SEM, n ¼ 3. Statistical
significance was assessed by calculating the p-value using one-way
ANOVA with Bonferroni correction in R, *p < 0.05.
incubated all three cell types with 10 mM RPS-1 for 3 hours,
followed by media exchange and treatment with 5 mM astemi-
zole and 30 mM K+ or with vehicle control in HEPES buffer for 1
hour. We observed an increase in the ratio of green/blue uo-
rescence in the HT29 cells compared to vehicle control and Hela
and A549 cell lines (Fig. 4). The astermizole-induced increases
in intracellular K+ levels were veried by ICP-MS (Fig. S5†).
Taken together, these data establish that RPS-1 is indeed able to
visualize both increases and decreases in intracellular K+ levels
by ratiometric uorescence imaging.
Fig. 5 Profiling relative resting intracellular K+ levels in various cancer
and non-cancerous (MCF-10A) cell lines by RPS-1 and ICP-MS. Data
given for (a) RPS-1 and (b) ICP-MS. Average fluorescence intensity
ratios are measured using ImageJ. Error bars denote SEM, n ¼ 3.
Statistical significance was assessed by calculating the p-value using
one-way ANOVA with Bonferroni correction in R, *p < 0.05.
measurements of bulk intracellular K+ levels validate the RPS-
1 assay by providing an independent measure of potassium
(Fig. 5 and S8†).
RPS-1 enables proling of intracellular K+ levels across a panel
of cancer cell lines
Concluding remarks
Aer establishing that RPS-1 is capable of monitoring dynamic
changes in intracellular K+ levels upon external stimulation, we To summarize, we have presented the design, synthesis, char-
sought to expand applications of RPS-1 to proling relative acterization, and biological applications of RPS-1, a rst-
intracellular K+ levels across different cell types. We became generation chemical probe for ratiometric imaging of intracel-
interested in comparative screening of K+ levels between cancer lular K+ pools. RPS-1 features a dual-uorophore approach to
and non-cancer cell models, as many types of tumors are known ratiometric potassium detection where the two uorophores
to exhibit elevated K+ channel expression and elevated K+ has operate independently to provide
a potassium-sensitive
been observed in some cases.18,21,67 We turned our attention to response in one dye with internal calibration from the other
comparing K+ levels in two different types of metastatic breast dye. RPS-1 is composed of two units: PS525 as a turn-on K+-
cancer cell lines, MDA-MB-468 and MDA-MB-231, relative to responsive uorophore with a lariat crown ether receptor
a normal breast cancer cell control line MCF-10A. Given the tailored to selective but low-affinity K+ binding, and Coumarin
ability of RPS-1 to reliably measure the relative intracellular K+ 343 as a K+-insensitive uorophore. When joined by ester
level in multiple types of cell lines, we also screened other linkages in the full RPS-1 probe, both components are taken up,
cancer cell lines, including U2OS (bone), RKO (colon), and PC-3 released, and trapped in cells aer cleavage by intracellular
(prostate), using HeLa as a comparison. Interestingly, RPS-1 esterases. Through the PS525 unit, RPS-1 can sensitively and
imaging reveals that the breast cancer cell lines MDA-MB-468 specically respond to intracellular K+ as well as report dynamic
and MDA-MB-231 possess higher basal levels of K+ compared changes in K+ levels upon stimulation with an external input.
to the normal breast cell line MCF-10A (Fig. 5). We speculate Compared to previous potassium probes such as PBFI, which
that the expansion of K+ pools is due to higher proliferation requires high-energy UV excitation, visible light excitation of
needs for metastatic tumor cells versus normal ones. In addi- RPS-1 through the PS525 chromophore minimizes background
tion, PC-3 and U2OS lines also exhibit higher resting levels of K+ autouorescence and photodamage. Moreover, due to its
over RKO and HeLa, showing that different cell types can indeed ratiometric response RPS-1 can be used to prole basal intra-
possess different basal K+ pools. Moreover, ICP-MS cellular K+ pools across a variety of different cell lines, revealing
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
that breast cancer cells possess higher resting levels than a non- CuCl2$2H2O (Baker & Adamson), CoCl2$6H2O (Sigma), MgCl2-
cancerous counterpart. We anticipate that RPS-1 and related $6H2O (Sigma), NaCl (Sigma), and FeCl2 (Sigma) were used.
uorescent probes will be of value in providing foundational
information about K+ physiology and pathology owing to the
broad importance of this abundant metal in biology, as well as
promote further investigations into design strategies for ratio-
Synthesis of RPS-1
metric detection of metals and other important biological
analytes.
4-(1,4,7,10,13-Pentaoxa-16-azacyclooctadecan-16-yl)-3-(2-
methoxyethoxy)benzaldehyde (2). POCl3 (3.72 mL, 40 mmol)
was added dropwise to a vigorously stirring portion of anhy-
drous DMF which was kept in an ice bath in a N2 atmosphere.
The resulting pale yellow solution was stirred at room temper-
ature for an additional 15 min. To this mixture, a DMF (2 mL)
solution of 16-(2-(2-methoxyethoxy)phenyl)-aza-18-crown-6 1
Experimental section
General methods
Reactions using moisture- or air-sensitive reagents were carried (ref. 42) (1.65 g, 4 mmol) was then slowly introduced and the
out in ame-dried glassware in an inert atmosphere of N2. resulting dark red solution was heated at 70 ꢁC for 3 h. The
Solvent was passed over activated alumina and stored over reaction was cooled to room temperature, and poured into an
˚
activated 3 A molecular sieves before use when dry solvent was an ice-cold satd NaHCO3 solution. The mixture was extracted
required. All other commercially purchased chemicals were used with 3 ꢂ 100 mL DCM/toluene (2 : 1) and dried over anhydrous
as received (without further purication). SiliCycle 60 F254 silica MgSO4. The solvent was evaporated in vacuo and the compound
gel pre-coated sheets (0.25 mm thick) were used for analytical was puried by silica gel ash column chromatography (FCC)
thin layer chromatography and visualized by uorescence using DCM : MeOH (19 : 1) as the eluent and aldehyde 2 was
quenching under UV light. Silica gel P60 (SiliCycle) was used for obtained as a dark red oil (820 mg, 44% yield) 1H NMR (400
1
column chromatography. H and 13C NMR NMR spectra were MHz, CDCl3) d (ppm): 9.48 (s, 1H), 7.38 (d, 2H, J ¼ 7.5 Hz), 7.17
collected at 298 K in CDCl3 or CD3OD (Cambridge Isotope (d, 2H, J ¼ 1.5 Hz), 6.47 (dd, 2H, J ¼ 7.5, 1.5 Hz), 4.31 (t, 2H, J ¼
ꢁ
Laboratories, Cambridge, MA) at 25 C using Bruker AVQ-400, 8.2 Hz), 3.77 (d, 2H, J ¼ 8.2 Hz), 3.60–3.52 (m, 24H), 3.41 (s, 3H).
AVB-400, AV-500, or AV-600 instruments at the College of 13C NMR (101 MHz, CDCl3) d (ppm): 191.01, 162.72, 148.11,
Chemistry NMR Facility at the University of California, Berkeley 127.08, 122.39, 115.44, 115.09, 72.21, 69.01, 59.79, 59.12. LCMS
or using a Bruker 900 at the QB3 Central California 900 MHz calcd for C22H35NO8 [M + H]+ 442.24, found 442.19.
NMR Facility. All chemical shis are reported in the standard
9-(4-(1,4,7,10,13-Pentaoxa-16-azacyclooctadecan-16-yl)-3-(2-
notation of d parts per million relative to the residual solvent methoxyethoxy)phenyl)-6-hydroxy-3H-xanthen-3-one (3). Alde-
peak at 7.26 (CDCl3) or 3.31 (CD3OD) for 1H and 77.16 (CDCl3) or hyde 2 (150 mg, 0.34 mmol, 1 equiv.) and resorcinol (82 mg,
49.00 (CD3OD) for 13C as an internal reference. Splitting patterns 0.74 mmol, 2.2 equiv.) were dissolved in 4.25 mL of triuoro-
are indicated as follows: br, broad; s, singlet; d, doublet; t, triplet; acetic acid (TFA) in a 35 mL pressure ask. The solution was
m, multiplet; dd, doublet of doublets. Low-resolution electro- then stirred at 115 ꢁC for 13 h and concentrated by rotary
spray mass spectral analyses were carried out using LC-MS evaporation to yield a crude red oil. The red oil was taken up in
(Agilent Technology 6130, Quadrupole LC/MS and Advion a silica column for purication by ash column chromatog-
Expression-L Compact Mass Spectrometer). High-resolution raphy (50% MeCN/toluene) to yield 3 (35 mg, 18% yield) as
1
mass spectral analyses (ESI-MS) were carried out at the College a bright yellow solid. H NMR (400 MHz, CDCl3) d (ppm): 7.33
of Chemistry Mass Spectrometry Facility at the University of (d, 2H, J ¼ 7.3 Hz), 6.84 (d, 2H, J ¼ 1.8 Hz), 6.47–6.11 (m, 7H),
California, Berkeley. Compound 1 was synthesized using the 4.31 (t, 2H, J ¼ 8.2 Hz), 3.77 (d, 2H, J ¼ 8.2 Hz), 3.60–3.52 (m,
published procedure.42 All aqueous solutions were prepared 24H), 3.41 (s, 3H). 13C NMR (101 MHz, CDCl3) d (ppm): 185.70,
using Milli-Q water, and all spectroscopic experiments were 162.72, 161.51, 155.92, 148.11, 127.08, 122.39, 115.44, 115.09,
carried out in 50 mM HEPES, pH 7.4, unless otherwise noted. All 95.73, 72.21, 69.01, 59.79, 59.12. LRMS calcd for C34H41N O10 [M
spectroscopic experiments were prepared using freshly prepared + H]+ 624.70, found 624.49.
aliquots. Absorption spectra were acquired using a Varian Cary
50 spectrophotometer, and uorescence spectra were acquired methoxyethoxy)phenyl)-3-oxo-3H-xanthen-6-yl
9-(4-(1,4,7,10,13-Pentaoxa-16-azacyclooctadecan-16-yl)-3-(2-
tri-
using a Photon Technology International Quanta Master 4 L- uoromethanesulfonate (4). 3 (414 mg, 1 mmol) was suspended
format scan spectro-uorometer equipped with an LPS-220B in MeCN (10 mL). N-Phenyl-bis(triuoromethane) sulfonimide
75 W xenon lamp and power supply, A-1010B lamp housing (1.071 g, 3 mmol) was then added and the solution was allowed
with an integrated igniter, switchable 814 photocounting/analog to stir for 3 h at room temperature. The solution was then
photomultiplier detection unit, and MD5020 motor driver. 1 cm evaporated and the crude compound was puried by column
ꢂ 1 cm quartz cuvettes (1.4 mL volume, Starna, capped) were chromatography (silica gel, 5–20% MeOH in DCM) to afford 4 as
used for obtaining absorption and uorescence spectra. For all a transparent light yellow oil (433 mg, 0.57 mmol, 57%) 1H NMR
uorescence measurements and K+ studies, aqueous solutions (400 MHz, CDCl3): d 7.20–7.12 (m, 2H), 6.89–6.30 (m, 6H), 6.21
of KCl (Sigma) were used. For metal selectivity studies, aqueous (d, 1H, J ¼ 1.6 Hz), 4.31 (t, 2H, J ¼ 6.4 Hz), 3.77 (t, 2H, J ¼ 6.4
metal solutions of MgCl2$4H2O (EMD Millipore), CaCl2$2H2O Hz), 3.64–3.54 (m, 24H), 3.35 (s, 3H). 13C NMR (101 MHz,
(EMD Millipore), NiCl2$6H2O (Sigma), ZnCl2 (Sigma), CDCl3): d 156.46, 149.93, 145.99, 134.71, 128.82, 125.07, 124.30,
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
123.97, 123.76, 115.57, 72.54, 70.42, 69.08, 61.84, 55.35, 49.43. with 10 mL portions of EtOAc, and the combined organic
LRMS calcd for C35H40F3NO12S [M + H]+ 755.76; found: 755.70. layers were evaporated. Purication by high-performance
tert-Butyl-4-(9-(4-(1,4,7,10,13-pentaoxa-16-azacyclooctade-
can-16-yl)-3-(2-methoxyethoxy)phenyl)-3-oxo-3H-xanthen-6-yl)
liquid chromatography (Agilent) on a C18 column using
a linear gradient of 20–100% CH3CN over 50 minutes gave
1
piperazine-1-carboxylate (5). Compound 4 (433 mg, 0.57 mmol), PS525 as a light-yellow powder (23 mg, 0.03 mmol, 48%). H
BINAP (124 mg, 0.20 mmol), and palladium acetate (11 mg, 0.05 NMR (400 MHz, CD3CN): d 7.35 (m, 2H), 6.99–6.37 (m, 7H),
mmol) were suspended in THF (5 mL) and stirred at room 6.28 (d, 1H, J ¼ 1.4 Hz), 4.31 (t, 2H, J ¼ 6.7 Hz), 3.77 (t, 2H, J ¼
temperature. 1-Boc-piperazine (279, 1.5 mmol) was then added 6.7 Hz), 3.64–3.54 (m, 27H), 3.37–3.29 (m, 4H), 3.22 (s, 2H),
and the solution was reuxed in a nitrogen atmosphere over- 2.51 (t, 4H, J ¼ 5.3 Hz). 13C NMR (101 MHz, CD3CN): d 180.91,
night. The reaction was then allowed to cool and ltered. The 176.99, 166.01, 147.32, 145.55, 141.72, 137.07, 133.21, 129.58,
ltrate was evaporated and subjected to silica gel chromatog- 123.97, 123.42, 119.03, 112.40, 110.36, 109.49, 107.48, 79.50,
raphy (40–100% EtOAc in hexanes) to yield 5 as a tan oil 70.02, 69.98, 69.22, 69.10, 55.27, 71.03, 55.12, 42.83, 42.40,
(214 mg, 0.26, 46%). 1H NMR (400 MHz, CDCl3): d 7.22 (m, 2H), 28.07. HRMS-ESI calculated for C41H53N3O11 [M + H+]:
6.82–6.30 (m, 7H), 6.21 (d, 1H, J ¼ 1.6 Hz), 4.31 (t, 2H, J ¼ 6.4 750.1757; found: 750.1764.
Hz), 3.77 (t, 2H, J ¼ 6.4 Hz), 3.64–3.54 (m, 24H), 3.35 (s, 3H),
Ratiometric Potassium Sensor-1, RPS-1. PS525 (21 mg, 0.03
3.32–3.29 (m, 8H), 1.41 (s, 9H). 13C NMR (101 MHz, CDCl3): mmol), compound 9 (ref. 63) (19 mg, 0.06 mmol), and TATU
d 164.25, 157.24, 152.89, 152.51, 141.87, 139.99, 137.24, 135.91, (32 mg, 0.10 mmol) were dissolved in 2 mL of dry DMF and
134.98, 132.44, 131.21, 130.74, 130.39, 124.57, 122.88, 67.21, stirred in a nitrogen atmosphere. DBU was then added to the
62.46, 56.49, 28.01, 20.85. LRMS calcd for C38H50N3O15 [M + H]+ solution. The reaction was kept at room temperature for 12 h.
791.46; found: 791.5.
Purication by high-performance liquid chromatography (Agi-
9-(4-(1,4,7,10,13-Pentaoxa-16-azacyclooctadecan-16-yl)-3-(2-
lent) was carried out on a C18 column using a linear gradient of
methoxyethoxy)phenyl)-6-(piperazin-1-yl)-3H-xanthen-3-one (6). 40–100% CH3CN over 50 min. Fractions containing the product
Compound 5 was dissolved in 5 mL of DCM. 5 mL of TFA was were combined, evaporated under vacuum and lyophilized to
1
added to the solution in a nitrogen atmosphere. The mixture give RPS-1 as a brown powder (11.7 mg, 0.014 mmol, 48%). H
was stirred for 5 h at ambient temperature. The resulting NMR (400 MHz, CD3CN): d 8.32 (s, 1H), 7.35 (m, 2H), 7.28 (s,
solution was evaporated under reduced pressure to yield a dark 1H), 6.99–6.37 (m, 7H), 6.28 (d, 1H, J ¼ 1.4 Hz), 4.31 (t, 2H, J ¼
brown oil crude product (198 mg). The crude product 6 was used 6.7 Hz), 4.13 (t, 2H, J ¼ 6.2 Hz), 3.97 (t, 2H, J ¼ 5.8 Hz), 3.77 (t,
for the following step without purication.
2H, J ¼ 6.7 Hz), 3.64–3.54 (m, 27H), 3.37–3.29 (m, 8H), 3.22 (s,
Methyl
2-(4-(9-(4-(1,4,7,10,13-pentaoxa-16-azacyclooctade- 2H), 2.79 (m, 4H), 2.51 (t, 4H, J ¼ 5.3 Hz), 1.81–1.60 (m, 8H) 13
C
can-16-yl)-3-(2-methoxyethoxy)phenyl)-3-oxo-3H-xanthen-6-yl)
NMR (101 MHz, CD3CN): d 180.93, 177.26, 166.42, 155.72,
piperazin-1-yl)acetate (7). Compound 6 from the previous 153.70, 147.65, 145.88, 141.95, 137.27, 136.07, 133.21, 124.21,
deprotection step was dissolved in 5 mL of dry MeCN. Methyl 123.71, 119.45, 112.76, 110.76, 109.81, 107.97, 107.79, 79.96,
bromoacetate (141 mL, 1.5 mmol) and sodium carbonate 70.36, 70.33, 69.58, 69.47, 59.54, 55.63, 55.47, 52.48, 51.23,
(742 mg, 7 mmol) were then added and the solution was 43.18, 42.73, 28.36, 26.17, 25.91, 23.01. HRMS-ESI calculated for
reuxed for 13 h. Aer 13 h, the mixture was cooled down to
ambient temperature and poured into water (50 mL) and then
extracted with ethyl acetate (3 ꢂ 50 mL). The combined organic
layers were dried over sodium sulfate, ltered and concentrated
via rotary evaporation. The crude residue was puried using
C
40H51N3O11 Na [M + Na+]: 771.1142; found: 771.1183.
Spectroscopic methods
Fluorescence turn-on responses to potassium. Two different
ash chromatography (silica gel, 0–20% MeOH in DCM) to yield 50 mM HEPES (pH 7.4) solutions, solution A and solution B
7 as a foamy brown solid (93 mg, 0.12 mmol) 1H NMR (400 MHz, were made. Solution A contained 200 mM Na+ and solution B
CD3CN): d 7.35 (m, 2H), 6.99–6.37 (m, 7H), 6.28 (d, 1H, J ¼ 1.4 contained 200 mM K+. Solution A and B were mixed in different
Hz), 4.31 (t, 2H, J ¼ 6.7 Hz), 3.77 (t, 2H, J ¼ 6.7 Hz), 3.64–3.54 (m, ratios to yield 1 mL nal buffer solution with 0, 5, 10, 20, 30, 40,
27H), 3.37–3.29 (m, 7H), 3.22 (s, 2H), 2.51 (t, 4H, J ¼ 5.3 Hz). 13
C
50, 70, 90, 110, 130, 150, 200 mM respectively. A 2 mM solution
NMR (101 MHz, CD3CN): d 166.42, 147.65, 145.88, 141.95, of PS525 was prepared by diluting 1 mM DMSO stock solution of
137.27, 133.71, 131.64, 124.21, 123.71, 119.45, 112.76, 110.76, PS525 with each buffer in a 1 : 500 ratio into a 1 cm ꢂ 1 cm
109.81, 107.97, 107.79, 79.96, 70.36, 70.33, 69.58, 69.47, 55.63, capped quartz cuvette. The probe solution was incubated at
55.47, 43.18, 42.73, 28.36. LC-MS calculated for C37H41BrN4- 37 ꢁC for 5 minutes. Emission spectra were collected for PS525
O14S2Na [M + Na+]: 763.11; found: 763.31.
(lex ¼ 525 nm, lem ¼ 545–700 nm).
Potassium Sensor 525, PS525, (8). Compound 7 (50 mg, 0.06
Metal selectivity experiments. A 2 mM solution of PS525 was
mmol) was suspended in 5 mL of THF. Then 5 mL of 1 M LiOH prepared by diluting a 1 mM DMSO stock solution of the probe
solution was added slowly to the above mixture while stirring. into 1 mL 50 mM HEPES (pH 7.4). 500 mL aliquots of this
Aer 1 h of stirring, the mixture had become homogeneous. solution were added to ten 1 cm ꢂ 1 cm capped quartz cuvettes
Aer 24 h, the solution was put in an ice bath and 2 mL of 3 M and then 500 mL of the metal of interest was added to the cuvette
HCl was added, and the mixture was allowed to warm to room to bring the concentration of transition metals to 10 mM and
temperature. Following the addition of 20 mL of saturated with the exception of Na+ at 5 mM. The emission spectrum was
aqueous NaCl solution, the mixture was extracted four times then recorded for each metal of interest solution.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
3 D. A. Baylor and J. G. Nicholls, J. Physiol., 1969, 203, 555–569.
4 A. A. Grace and B. S. Bunney, J. Neurosci., 1984, 4, 2877–2890.
5 D. L. Fortin, M. R. Banghart, T. W. Dunn, K. Borges,
D. A. Wagenaar, Q. Gaudry, M. H. Karakossian, T. S. Otis,
W. B. Kristan, D. Trauner and R. H. Kramer, Nat. Methods,
2008, 5, 331–338.
6 P. M. Hudgins and G. B. Weiss, J. Pharmacol. Exp. Ther., 1968,
159, 91–97.
7 C. Juel, Puegers Arch., 1986, 406, 458–463.
8 M. J. Bia and R. A. DeFronzo, Am. J. Physiol., 1981, 240, F257–
F268.
Preparation and staining of cell cultures. All cells were
maintained at exponential growth as a monolayer in Dulbeco's
Modied Eagle Medium (DMEM, Invitrogen). Cells were sup-
plemented with 10% fetal bovine serum (FBS, Hyclone) and
incubated at 37 ꢁC in 5% CO2. Cells were plated in phenol red-
free medium on borosilicate chamber slides (Nunc) and allowed
to grow to 60% conuence 1 day prior to imaging. Cells were
stained by incubating a 10 mM solution of RPS-1 in DMEM for
2 h. For the valinomycin-induced K+ efflux experiments, the
DMEM media was exchanged out for phosphate buffered metal
solution (PBS, pH 7.4, M ¼ Na or K such that [Na + K]tot ¼ 160
mM) containing 10 mM valinomycin (Invitrogen). The cells were
allowed to equilibrate in the new media for 1 min prior to
imaging.
´
9 E. Feraille and A. Doucet, Physiol. Rev., 2001, 81, 345–418.
10 N. M. Storey, J. P. O'Bryan and D. L. Armstrong, Curr. Biol.,
2002, 12, 27–33.
11 J. M. Burnell, B. H. Scribner, B. T. Uyeno and M. F. Villamil, J.
Clin. Invest., 1956, 35, 935–939.
12 C. Tian, R. Zhu, L. Zhu, T. Qiu, Z. Cao and T. Kang, Chem.
Biol. Drug Des., 2014, 83, 1–26.
13 L. Ge, N. T. Hoa, Z. Wilson, G. Arismendi-Morillo, X. T. Kong,
R. B. Tajhya, C. Beeton and M. R. Jadus, Int.
Immunopharmacol., 2014, 22, 427–443.
ICP-MS. Cells were then rinsed twice with ice-cold potassium
free buffer (150 mM Na+ in 50 mM HEPES buffer, pH ¼ 7.4),
followed by the addition of 215 mL of concentrated nitric acid
(BDH Aristar Ultra). The plates were sealed with Paralm and
incubated on a shaker overnight. Samples (150 mL) were further
diluted in 2 mL of 2% nitric acid (made freshly from concen-
trated nitric acid and Milli-Q water) in 15 mL tubes (Sarstedt)
and analyzed on a Thermo Fisher iCAP Qc ICP mass spec-
trometer in kinetic energy discrimination (KED) mode against
14 A. Felipe, R. Vicente, N. Villalonga, M. Roura-Ferrer,
´
´
´
R. Martınez-Marmol, L. Sole, J. C. Ferreres and E. Condom,
Cancer Detect. Prev., 2006, 30, 375–385.
15 H. Wulff, N. A. Castle and L. A. Pardo, Nat. Rev. Drug
Discovery, 2009, 8, 982–1001.
16 S. A. Salyer, J. R. Olberding, A. A. Distler, E. D. Lederer,
B. J. Clark, N. A. Delamere and S. J. Khundmiri, Biochim.
Biophys. Acta, Mol. Basis Dis., 2013, 1832, 1734–1743.
a
standard curve of known potassium and phosphorus
concentrations (CMS-5, Inorganic Ventures), with Ga (20 mg Lꢀ1
,
Inorganic Ventures) as an internal standard. Each experiment
was carried out twice, and each condition was repeated in at
least triplicate.
Confocal microscopy imaging. Microscopy experiments were
performed using a Zeiss laser scanning microscope (model 710)
equipped with a 40ꢂ water objective lens and 2010 Zeiss so-
¨
17 D. Urrego, A. P. Tomczak, F. Zahed, W. Stuhmer, L. A. Pardo
and W. Stuehmer, Philos. Trans. R. Soc., B, 2014, 2014, 1–9.
18 X. Huang and L. Y. Jan, J. Cell Biol., 2014, 206, 151–162.
19 L. Leanza, E. Venturini, S. Kadow, A. Carpinteiro, E. Gulbins
and K. A. Becker, Cell Calcium, 2015, 58, 1–8.
ꢁ
ware (Carl Zeiss). Cells were imaged in a 37 C stage incubator
connected to 5% CO2 gas inlet to buffer the media. The focal
plane with the highest uorescence was selected for each eld
of cells. Fields of cells were sampled from the bright eld
images only. Replicating experiments were conducted with
different plates of cells.
20 L. Leanza, M. Romio, K. A. Becker, M. Azzolini, L. Trentin,
`
A. Manago, E. Venturini, A. Zaccagnino, A. Mattarei,
L. Carraretto, A. Urbani, S. Kadow, L. Biasutto, V. Martini,
F. Severin, R. Peruzzo, V. Trimarco, J. H. Egberts,
C. Hauser, A. Visentin, G. Semenzato, H. Kalthoff,
M. Zoratti, E. Gulbins, C. Paradisi and I. Szabo, Cancer
Cell, 2017, 31, 516–531.e10.
Conflicts of interest
There are no conicts to declare.
21 E. K. Breuer, D. Fukushiro-Lopes, A. Dalheim, M. Burnette,
J. Zartman, S. Kaja, C. Wells, L. Campo, K. J. Curtis,
R. Romero-Moreno, L. E. Littlepage, G. L. Niebur,
K. Hoskins, M. I. Nishimura and S. Gentile, Cell Death Dis.,
2019, 10, 1–15.
22 C. D. Bortner, F. M. Hughes and J. A. Cidlowski, A Primary
Role for K and Na Efflux in the Activation of Apoptosis, J.
Biol. Chem., 1997, 272, 32436–32442.
Acknowledgements
This work was supported by the Human Frontiers Science
Program Organization (RGP0052/2015) and NIH (GM 79465). T.
C. D. was partially supported by a Chemistry–Biology Interface
Training Grant from the NIH (T32 GM066698). We thank
members of the Chang lab for feedback on the manuscript.
Z. W. acknowledges support from Suzhou Industrial Park.
23 F. M. Hughes, C. D. Bortner, G. D. Purdy and J. A. Cidlowski,
J. Biol. Chem., 1997, 272, 30567–30576.
ˇ
˚ˇ
24 U. Fiedler and J. Ruzicka, Anal. Chim. Acta, 1973, 67, 179–
Notes and references
193.
´
1 M. C. D'Adamo, L. Catacuzzeno, G. Di Giovanni, 25 M. Jospin, M. C. Mariol, L. Segalat and B. Allard, J. Physiol.,
F. Franciolini and M. Pessia, Front. Cell. Neurosci., 2013, 7,
1–21.
2 B. F. Palmer, Clin. J. Am. Soc. Nephrol., 2014, 10, 1050–1060.
2002, 544, 373–384.
26 H. Bischof, M. Rehberg, S. Stryeck, K. Artinger, E. Eroglu,
M. Waldeck-Weiermair, B. Gottschalk, R. Rost, A. T. Deak,
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
T. Niedrist, N. Vujic, H. Lindermuth, R. Prassl, B. Pelzmann, 47 P. Padmawar, X. Yao, O. Bloch, G. T. Manley and
K. Groschner, D. Kratky, K. Eller, A. R. Rosenkranz, T. Madl, A. S. Verkman, Nat. Methods, 2005, 2, 825–827.
N. Plesnila, W. F. Graier and R. Malli, Nat. Commun., 2017, 8, 48 W. Namkung, P. Padmawar, A. D. Mills and A. S. Verkman, J.
1–11. Am. Chem. Soc., 2008, 130, 7794–7795.
27 T. T. Ruckh, C. G. Skipwith, W. Chang, A. W. Senko, 49 R. D. Carpenter and A. S. Verkman, Org. Lett., 2010, 12, 1160–
V. Bulovic, P. O. Anikeeva and H. A. Clark, ACS Nano, 2016,
10, 4020–4030.
28 C. H. Lee, J. Folz, W. Zhang, J. Jo, J. W. Y. Tan, X. Wang and
R. Kopelman, Anal. Chem., 2017, 89, 7943–7949.
1163.
50 R. D. Carpenter and A. S. Verkman, Eur. J. Org. Chem., 2011,
2011, 1242–1248.
51 C. Aer, Nat. Methods, 2005, 2, 5–8.
29 R. Wang, X. Du, Y. Wu, J. Zhai and X. Xie, ACS Sens., 2018, 3, 52 J. Yu, L. Zhang, X. Xu and S. Liu, Anal. Chem., 2014, 86,
2408–2414. 10741–10748.
30 Y. Shen, S. Y. Wu, V. Rancic, A. Aggarwal, Y. Qian, 53 J. Huang, L. Ying, X. Yang, Y. Yang, K. Quan, H. Wang,
S. I. Miyashita, K. Ballanyi, R. E. Campbell and M. Dong,
Commun. Biol., 2019, 2, 1–10.
N. Xie, M. Ou, Q. Zhou and K. Wang, Anal. Chem., 2015,
87, 8724–8731.
31 M. P. Jewell, M. D. Greer, A. L. Dailey and K. J. Cash, ACS 54 G. Grynkiewicz, M. Poenie and R. Y. Tsien, J. Biol. Chem.,
Sens., 2020, 5, 474–480.
1985, 260, 3440–3450.
32 R. Y. Tsien, Fluorescent Probes of Cell Signaling, 1989.
33 D. W. Domaille, E. L. Que and C. J. Chang, Nat. Chem. Biol.,
2008, 4, 168–175.
34 K. P. Carter, A. M. Young and A. E. Palmer, Chem. Rev., 2014,
114, 4564–4601.
55 A. T. Aron, M. O. Loehr, J. Bogena and C. J. Chang, J. Am.
Chem. Soc., 2016, 138, 14338–14346.
56 C. Y. S. Chung, J. M. Posimo, S. Lee, T. Tsang, J. M. Davis,
D. C. Brady and C. J. Chang, Proc. Natl. Acad. Sci. U. S. A.,
2019, 116, 18285–18294.
35 S. C. Dodani, A. Firl, J. Chan, C. I. Nam, A. T. Aron, 57 C. C. Woodroofe and S. J. Lippard, J. Am. Chem. Soc., 2003,
C. S. Onak, K. M. Ramos-Torres, J. Paek, C. M. Webster, 125, 11458–11459.
M. B. Feller and C. J. Chang, Proc. Natl. Acad. Sci. U. S. A., 58 R. A. Schultz, B. D. White, D. M. Dishone, K. A. Arnold and
2014, 111, 16280–16285. G. W. Gokel, J. Am. Chem. Soc., 1985, 107, 6659–6668.
36 C. M. Ackerman, S. Lee and C. J. Chang, Anal. Chem., 2017, 59 G. W. Gokel, L. J. Barbour, R. Ferdani and J. Hu, Acc. Chem.
89, 22–41. Res., 2002, 35, 878–886.
37 J. L. Kolanowski, F. Liu and E. J. New, Chem. Soc. Rev., 2018, 60 A. P. De Silva, H. Q. N. Gunaratne, T. Gunnlaugsson,
47, 195–208.
A. J. M. Huxley, C. P. McCoy, J. T. Rademacher and
T. E. Rice, Chem. Rev., 1997, 97, 1515–1566.
61 J. E. Whitaker, R. P. Haugland, D. Ryan, P. C. Hewitt,
R. P. Haugland and F. G. Prendergast, Anal. Biochem.,
1992, 207, 267–279.
38 H. He, M. A. Mortellaro, M. J. P. Leiner, R. J. Fraatz and
J. K. Tusa, J. Am. Chem. Soc., 2003, 125, 1468–1469.
39 F. He, Y. Tang, S. Wang, Y. Li and D. Zhu, J. Am. Chem. Soc.,
2005, 127, 12343–12346.
40 S. Nagatoishi, T. Nojima, B. Juskowiak and S. Takenaka, 62 T. Xiao, C. M. Ackerman, E. C. Carroll, S. Jia, A. Hoagland,
Angew. Chem., 2005, 117, 5195–5198.
41 X. Fan, H. Li, J. Zhao, F. Lin, L. Zhang, Y. Zhang and S. Yao,
Talanta, 2012, 89, 57–62.
J. Chan, B. Thai, C. S. Liu, E. Y. Isacoff and C. J. Chang,
Nat. Chem. Biol., 2018, 14, 655–663.
63 C. C. Woodroofe, A. C. Won and S. J. Lippard, Inorg. Chem.,
2005, 44, 3112–3120.
¨
42 S. Ast, T. Schwarze, H. Muller, A. Sukhanov, S. Michaelis,
¨
¨
¨
J. Wegener, O. S. Woleis, T. Korzdorfer, A. Durkop and 64 E. Lastraioli, L. Guasti, O. Crociani, S. Polvani, G. Hofmann,
H. J. Holdt, Chem.–Eur. J., 2013, 19, 14911–14917.
43 X. Kong, F. Su, L. Zhang, J. Yaron, F. Lee, Z. Shi, Y. Tian and
D. R. Meldrum, Angew. Chem., Int. Ed., 2015, 54, 12053–
12057.
44 B. Sui, X. Yue, M. G. Tichy, T. Liu and K. D. Beleld, Eur. J.
Org. Chem., 2015, 2015, 1189–1192.
H. Witchel, L. Bencini, M. Calistri, L. Messerini, M. Scatizzi,
R. Moretti, E. Wanke, M. Olivotto, G. Mugnai and
A. Arcangeli, Cancer Res., 2004, 606–611.
65 S. Z. Chen, M. Jiang and Y. S. Zhen, Cancer Chemother.
Pharmacol., 2005, 56, 212–220.
66 Z. Zhou, V. R. Vorperian, Q. Gong, S. Zhang and
C. T. January, J. Cardiovasc. Electrophysiol., 1999, 10, 836–
843.
45 G. Song, R. Sun, J. Du, M. Chen and Y. Tian, Chem. Commun.,
2017, 53, 5602–5605.
46 K. Meuwis, N. Boens, F. C. De Schryver, J. Gallay and 67 I. L. Camero, N. K. R. Smith, T. B. Pool and R. L. Sparks,
M. Vincent, Biophys. J., 1995, 68, 2469–2473. Cancer Res., 1980, 40, 1493–1500.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020