Electrophysiology is the gold standard for investigating ion channel activity. However, there are plenty of alternative approaches, including optical methods. Here, we describe a method to monitor the activity of the leucine-rich repeat containing 8 channel (LRRC8)-formed anion channels using an inter-subunit Förster resonance energy transfer (FRET)-based method.
Members of the LRRC8 protein family form heteromeric ion and osmolyte channels with roles in numerous physiological processes. As volume-regulated anion channels (VRACs)/volume-sensitive outwardly rectifying channels (VSORs), they are activated upon osmotic cell swelling and mediate the extrusion of chloride and organic osmolytes, leading to the efflux of water and hence cell shrinkage. Beyond their role in osmotic volume regulation, VRACs have been implicated in cellular processes such as differentiation, migration, and apoptosis. Through their effect on membrane potential and their transport of various signaling molecules, leucine-rich repeat containing 8 (LRRC8) channels play roles in neuron-glia communication, insulin secretion, and immune response. The activation mechanism has remained elusive. LRRC8 channels, like other ion channels, are typically studied using electrophysiological methods. Here, we describe a method to detect LRRC8 channel activation by measuring intra-complex sensitized-emission Förster resonance energy transfer (SE-FRET) between fluorescent proteins fused to the C-terminal leucine-rich repeat domains of LRRC8 subunits. This method offers the possibility to study channel activation in situ without exchange of the cytosolic environment and during processes such as cell differentiation and apoptosis.
Ion channels comprised of heteromers of leucine-rich repeat containing 8 (LRRC8) family proteins are found throughout vertebrate cells, participating in a wide range of physiological functions1,2. These LRRC8 channels, first identified as volume-regulated anion channels (VRACs) or volume-sensitive outwardly rectifying channels (VSOR), play a crucial role in cellular regulatory volume decrease3,4. They facilitate the expulsion of chloride ions and organic osmolytes, which is followed by water efflux in response to osmotic swelling. Beyond their role in osmotic stress response, their role in cellular volume regulation has been linked to cell proliferation and migration, apoptosis, spermiogenesis, and epithelial integrity5,6,7. Alteration of the membrane potential upon LRRC8/VRAC activation has been shown to contribute to myotube differentiation8 and insulin secretion by pancreatic β-cells9,10,11. Furthermore, LRRC8 channels conduct a variety of organic osmolytes such as purinergic signaling molecules ATP and cGAMP or the excitatory amino acid glutamate, placing these channels in cell-cell communication in the immune system or glia-neuron interaction12,13,14,15,16,17,18,19,20,21,22. Even xenobiotics, such as the dye fluorescein, the antibiotic blasticidin S or the anticancer drug cisplatin, are conducted by LRRC8 channels23,24,25.
There are numerous reports on the signal transduction leading to LRRC8/VRAC activation26,27,28. However, the mechanism remains unclear, and the literature presents a broad range of potential mechanisms that could depend on the specific physiological process. These include changes in cytosolic ion strength, interaction with the cytoskeleton, membrane composition, G proteins, the redox state, and phosphorylation cascades2,27,29,30,31.
LRRC8/VRAC channels contain LRRC8A as an essential subunit3,4 that must heteromerize with at least one of its paralogues LRRC8B-E to form physiologically functional channels4,14,32. The subunit composition determines biophysical properties of the channel, such as rectification and depolarization-dependent inactivation4,29,32,33,34, substrate specificity15,17,20,21,24,35, and some activation pathways36,37. Cryo-electron microscopy (cryo-EM) structures show that LRRC8A homomers, as well as heteromers, assemble as hexamers38,39,40, while LRRC8A/LRRC8C chimeras that form functional channels are heptamers41. The N-terminal part of all LRRC8 proteins comprises four transmembrane helices, and the C-terminal part contains a domain with leucine-rich repeats (LRRD). The available LRRC8 complex structures provide evidence that the LRRDs, which stretch into the cytosol3,4,23, may undergo conformational rearrangements during channel gating34,42,43. This notion is corroborated by the finding that C-terminal fusion of fluorescent proteins results in basal channel activity14 and that binding of nanobodies to the domains can modulate channel activity44. Moreover, conformational alterations of the C-termini were shown by intra-complex Förster resonance energy transfer (FRET)45.
The most common method to study ion channel activity is electrophysiological measurements46, which were extensively applied in the investigation of VRACs before their molecular identification47. However, there are various additional ways to monitor VRAC activity indirectly, including the measurement of its conducted substrates -halide ions or organic osmolytes- or its effect on cell volume48. In fact, the identification of LRRC8 proteins as VRAC relied on an assay based on the quenching of a halide-sensitive fluorescent protein49 by iodide entering the cell through activated VRACs3,4. Another method to monitor LRRC8/VRAC channel activity makes use of the movement of the cytosolic domains which can be observed, as in other ion channels50,51,52,53, by changes in FRET45. To this end, fluorescent proteins that serve as FRET pairs, such as cyan-fluorescent protein (CFP)/mCerulean3 as donor and yellow-fluorescent protein (YFP)/mVenus as acceptor, were fused to the C-termini of the LRRC8 proteins (Figure 1). Intra-complex FRET between LRRC8 subunits was shown by acceptor photobleaching experiments45. Avoiding the destructive photobleaching method, FRET changes over time were monitored by sensitized-emission FRET (SE-FRET), where basically the sensitized emission of the acceptor upon excitation of the donor due to the overlap of the emission spectrum of the donor with the excitation spectrum of the acceptor is measured. Application of extracellular hypotonicity, a stimulus for LRRC8/VRAC activation, resulted in a reversible reduction in SE-FRET intensity45. Importantly, simultaneous whole-cell patch-clamp measurements and FRET monitoring during hypotonic treatment showed that this reduction in FRET indeed mirrored LRRC8/VRAC activation45. This method, which avoids disrupting the plasma membrane or altering the intracellular environment by pipette solution, offers an alternative for monitoring LRRC8/VRAC activity. It is particularly useful in physiological settings where maintaining the native cytosol is crucial, subcellular resolution is necessary, or prolonged observation of channel activity is required.
Here, we present a protocol to study LRRC8/VRAC with such a FRET-based read-out. The protocol depicts how to handle and transfect cells, acquire sample and control images, analyze the data, and calculate sensitized emission FRET (SE-FRET) values.
Figure 1: Schematic of the LRRC8 FRET pair system. mCerulean3 is shown in cyan, and mVenus is shown in yellow. Following VRAC opening, the distance (and/or the spatial orientation) between the fluorophores changes, resulting in a reduced energy transfer between donor (Don) and acceptor (Acc) and, in turn, lowering the observed FRET. Created with BioRender.com. Please click here to view a larger version of this figure.
1. Preparation of buffer and reagents
2. Growth of adherent cells on glass bottom dishes
3. Cell transfection
NOTE: Here, FuGENE was used as a transfection reagent. Other transfection reagents and methods are also applicable. Optimal ratios of plasmid DNA (pDNA) to transfection reagents and time for optimal expression for each POI and cell model need to be assessed empirically. Here, 2 µg of total pDNA per 35 mm dish was used. FRET donor and acceptor constructs were used in a 1:1 ratio, and the pDNA-to-transfection reagent ratio was 1:4 (Table 1).
Condition | Construct (s) | Sample | Used for |
1 | LRRC8A-mCerulean | Donor construct only | determine correction factor β |
2 | LRRC8E-mVenus | Acceptor construct only | determine correction factor γ |
3 | LRRC8A-mCerulean and LRRC8E-mVenus | FRET pair | SE-FRET quantification |
Table 1: Example of conditions necessary for a typical SE-FRET experiment to measure LRRC8/VRAC activity of a channel composed of the LRRC8A subunit fused to the donor (mCerulean3) and LRRC8E subunit fused to the acceptor (mVenus) fluorophore.
4. Image acquisition for correction factor determination
NOTE: There is a bleed-through of the donor emission into the detected emission of the acceptor during FRET. Moreover, there is cross-excitation of the acceptor fluorophore by the donor excitation wavelength. These processes must be compensated for during the calculation of the SE-FRET. To this end, correction factors are determined in cells expressing only the FRET donor or acceptor 24 h after transfection. Here, the imaging was carried out on a Leica THUNDER Imager equipped with a Leica LED8 lamp, the filter cube CYR71010, an HC PL APO 63x/1.40 OIL objective, long pass filter for 460/80 and 553/70 and a Leica DFC9000GTC camera. Experiments were performed without environmental control but in the presence of HEPES in the imaging buffers to stabilize the pH. For long-term observation/measurements, it is advised to use an environmental control system. For analysis, SE-FRET is calculated from the captured raw images. This can be done simultaneously during acquisition or afterward. Here, the Leica LAS X software with the SE-FRET plugin was used to simplify the experimental procedure regarding the calculation of the correction factors and visualizing the SE-FRET value changes in real-time during image acquisition. For post-acquisition, correction factors and SE-FRET can be determined with other software packages (e.g., FIJI) after raw data acquisition according to the protocol provided below.
Excitation of | Emission of | Channel name | LED line | Filter cube | Long-pass filter |
Donor | Donor | DD | 440 nm | CYR71010 | 460/80 nm |
Donor | Acceptor | DA | 440 nm | CYR71010 | 535/70 nm |
Acceptor | Acceptor | AA | 510 nm | CYR71010 | 535/70 nm |
Table 2: Summary of channels required for SE-FRET experiments.
Figure 2: Representative fluorescence images of samples used to calculate the correction factors β and γ necessary to determine the VRAC activity of a channel composed of LRRC8A subunit fused to the donor mCerulean3 (mCer) and LRRC8E subunit fused to the acceptor mVenus (mVen) fluorophore by SE-FRET measurements. (A,B) Detection of the donor/donor DD, acceptor/acceptor AA, and donor/acceptor DA channel in HeLa cells only expressing the donor (A) LRRC8A-mCer or the acceptor (B) LRRC8E-mVen. (C) Detection of DD, AA, and DA channels in HeLa cells co-transfected with the donor and acceptor pair LRRC8A-mCer and LRRC8E-mVen. Panels a-i show images taken in the donor detection channel (excitation of the donor and detection of the donor signal; DD; a, d and g), the acceptor detection channel (excitation of the acceptor and detection of the acceptor signal; AA; b, e, and h ) and the FRET signal detection channel (excitation of the donor and detection of the acceptor signal; DA; c, f, and i). Panel j is the overlay of panels g and h. DD channel is shown in green and AA channel in magenta. Scale bar = 10 µm. Please click here to view a larger version of this figure.
5. Time-lapse imaging for SE-FRET quantification
Figure 3: Representative fluorescence images and SE-FRET quantification. (A) Representative fluorescence images and the apparent SE-FRET of the first time point of a time laps experiment to measure the VRAC activity by SE-FRET of a channel composed of LRRC8A and LRRC8E subunits depending on the tonicity. Scale bar = 10 µm. The same cells are shown in Figure 2C. Panels a–i show the detection of the DD, AA, and DA channels and the calculated apparent SE-FRET. White outlines represent the ROIs (cells i-iii in panel d) used to measure the mean signal intensities in DD, AA and DA and the apparent SE-FRET image. (B) Quantification of the SE-FRET values over time. Sequence of conditions was 12 cycles of isotonic imaging buffer (baseline) followed by 15 cycles of hypotonic and 15 cycles of hypertonic imaging buffer. The raw mean SE-FRET value of each ROI (cells i-iii) and time point was normalized to the mean of the baseline (isotonic) value for the corresponding ROI. Please click here to view a larger version of this figure.
With this FRET-based method, LRRC8/VRAC activity can be monitored during osmotic stimulation, and the reduction in SE-FRET correlates with the degree of extracellular hypotonicity45. Representative results for hypotonicity-induced channel activation are also shown here (Figure 3 and Figure 4). In addition, LRRC8/VRAC activation by different isosmotic stimuli, such as manipulation of diacylglycerol signaling45 or during myocyte activation56, can be observed.
Figure 4: SE-FRET traces. (A) Apparent SE-FRET traces from 5 independent experiments. Data represent mean ± SD of N = 2 to N = 7 cells per field of view (FOV). (B) Mean ± SD of all cells (N = 31 cells). Please click here to view a larger version of this figure.
As LRRC8/VRAC channels are also involved in apoptotic volume decrease24,57, observing channel activity upon induction of apoptosis would be another application for the method described here. Accordingly, the protocol to monitor SE-FRET in HeLa cells expressing LRRC8A-mCerulean3 and LRRC8E-mVenus was carried out while applying death receptor-mediated apoptosis-inducing drugs. Tumor necrosis factor (TNF)-α and cycloheximide (CHX) were previously shown to evoke VRAC currents within a few minutes58. After the addition of 2 ng/mL TNF-α and 1 µg/mL CHX in isotonic buffer, there was a robust decrease in SE-FRET (Figure 5). Replacing the buffer by a hypertonic medium, even though containing TNF-α and CHX, SE-FRET values recovered close to baseline (Figure 5A), corresponding to diminishing VRAC currents in hypertonic bath solution during treatment with apoptosis-inducing Fas ligand58. Treatment of the cells with DMSO, the solvent for TNF-α and CHX, did not result in SE-FRET reduction. TNF-α + CHX did not affect SE-FRET of CFP-18aa-YFP, an EYFP and ECFP tandem construct as FRET control59, demonstrating the specificity for LRRC8/VRAC (Figure 5B).
Figure 5: Isosmotic VRAC activation by death receptor-mediated apoptosis. (A) Normalized SE-FRET values from LRRC8A-mCer/LRRC8E-mVen-expressing HeLa cells (n = 8 dishes, 23 cells) over time. After 15 cycles in isotonic imaging buffer (baseline), the bath solution was replaced by isotonic buffer supplemented with 2 ng/mL TNF-α and 1 µg/mL cycloheximide (CHX) for 30 cycles followed by 20 cycles hypertonic imaging buffer with TNF-α and CHX. The raw mean SE-FRET value of each ROI and time point was normalized to the mean of the baseline (isotonic) value for the respective ROI. (B) Quantification of normalized SE-FRET values of LRRC8A/E-expressing HeLa cells as in A with isotonic solution containing DMSO as vehicle control for CHX (n = 5 dishes, 12 cells) or isotonic buffer containing apoptosis inducers as in A (n = 8 dishes, 23 cells), or of HeLa cells expressing CFP-18aa-YFP with apoptosis inducers (n = 3 dishes, 9 cells). Data represent the mean of the last 10 time points in the respective buffer of individual cells (symbols) and mean of all cells ± SD; ** p < 0.01 ordinary one-way ANOVA followed by Tukey´s multiple comparison post-hoc test. Please click here to view a larger version of this figure.
FRET microscopy is a well-established, widely used technique to study the interaction between proteins. Hence, FRET-based methods can be applied in many laboratories of variable expertise. Conformational rearrangements during gating have been monitored for a broad range of ion channels using FRET-based assays (for examples, see references34,50,51,52,53,60,61,62,63,64,65,66,67), in some cases combined with electrophysiology in patch-clamp fluorometry68,69,70,71. FRET can be used to study the structure-function relationships of these ion channels or to monitor their activity independent of ion transport. The method presented here can have clear advantages over electrophysiology as it allows monitoring of the activity of LRRC8/VRAC channels in situ.
Critical steps in the protocol include plating the cells to reach optimal confluency for transfection and imaging, which ideally facilitates easy cell distinction for later analysis. Effective co-transfection of the different subunits is crucial for correct subcellular localization; for example, an excess of the non-LRRC8A subunit will lead to enhanced endoplasmic reticulum (ER) localization4. Therefore, plasmid ratios may need to be adjusted. Depending on the system, newly generated FRET pairs should be verified, e.g., by acceptor bleaching. Binning and exposure time must be balanced against each other to enable optimal temporal and spatial resolution for the research question. Binning enables shorter exposure times and hence reduces potential bleaching of the FRET sensor while decreasing spatial resolution. Therefore, if the experimental setup requires, e.g., subcellular discrimination of LRRC8/VRAC activity, binning should be avoided. The research question equally determines the number and interval of cycles in a time-lapse series. The interval is only relevant if the kinetics of the FRET changes (and hence LRRC8/VRAC activation/inactivation) are required; otherwise, simple "before-and-after" recordings can also be performed. The length of the experiment depends on the physiological process. Ideally, LRRC8/VRAC activity upon stimuli should be monitored until SE-FRET has stabilized. These factors can be determined in pilot experiments. Correction factors to calculate the real SE-FRET signal must be determined for all conditions. Incorrectly determined correction factors may lead to an over- or underestimation of the SE-FRET intensities. Lastly, after establishing a stable baseline, the time interval between images has to be short enough to capture the physiological process of interest.
The method bears some limitations. One of them is that changes in inter-LRRC8 FRET intensities while reflecting movements of the LRRDs, do not necessarily correspond to ion or osmolyte transport through the pore. This is clear from the FRET changes observed with LRRC8A homomers45 despite their minimal currents4,32,72. Pore blockers of LRRC8/VRAC channels may not affect the FRET signal, rendering this method unsuitable for the search for specific channel modulators. Moreover, the expression levels of the overexpressed LRRC8 proteins could affect the physiological processes that are observed, especially as the C-terminally tagged LRRC8 proteins display basal activity14.
An aspect that can be considered a limitation or an advantage depending on the particular research question is that in this method, only the ectopically expressed LRRC8 subunits are selectively measured. So, background levels of endogenous proteins hardly interfere with the measurements. On the other hand, the overexpressed proteins may not behave like the endogenous LRRC8 channels with potentially different subunit composition and stoichiometry. For example, various stimuli such as oxidation may have opposing modulatory effects on differently composed LRRC8 channels36. By altering the ratios between co-expressed subunits, their stoichiometry, and overall ion conductance can be adjusted14,73, but their native composition, with likely often more than two paralogs within one complex21, is not clear and may vary between cell types74,75,76. Furthermore, the fusion of fluorescent proteins to the cytosolic C-termini of LRRC8 proteins was shown to increase basal LRRC8/VRAC channel activity in Xenopus oocytes14, likely because the large tags modulate the conformation of the LRRDs, which may govern channel opening14,44,45. Therefore, the size of the fluorescent proteins, the linker, and their orientation may not only affect FRET efficiency but also channel activity. However, importantly, VRAC channels of LRRC8 proteins fused with fluorescent proteins remained responsive to hypotonic stimulation14, enabling their use as FRET sensors45.
Advantages of this non-invasive method to monitor LRRC8/VRAC channel activity by light microscopy compared to other methods comprise: (i) It allows observing LRRC8/VRAC within cells or compartments typically inaccessible for electrophysiology. This includes intracellular organelles on which LRRC8 complexes can be found or targeted to45,77,78. (ii) The cytosolic composition remains unaltered by the method, whereas during whole-cell patch-clamp measurements, the cytosol is largely replaced by pipette solution, which may affect signaling pathways as observed with phorbol-12-myristate-13-acetate (PMA)-induced LRRC8/VRAC activation45. (iii) It offers the possibility to observe LRRC8/VRAC activation with subcellular resolution, such as distinguishing activity at the leading and trailing edges during cell migration, where -restricted to confined spaces- VRAC has been implicated79,80. (iv) It enables continuous monitoring of LRRC8/VRAC activity during extended physiological processes such as myocyte differentiation56.
While there are limitations and challenges with this method, it holds promise for further exploration, including potential applications in animal models. In combination with other methods to study this ion and osmolyte channel family, this FRET-based assay may contribute significantly to unraveling activation mechanisms and exploring the diverse physiological functions of LRRC8 channels in their native environments.
The authors have nothing to disclose.
We thank C.F. Kaminski for the kind gift of the plasmid encoding the CFP-18aa-YFP construct, A. Klemmer for technical assistance, and all current and former members of the Stauber laboratory who contributed to the development of this method.
0.05% Trypsin-EDTA | gibco | 25300-054 | |
Camera DFC9000GTC | Leica | 11547007 | |
CFP-18aa-YFP | N/A | N/A | Elder et al. 2009 PMCID: PMC2706461; Gift from C.F. Kaminski (University of Cambridge, UK) |
Cycloheximide (CHX) | Sigma-Aldrich | 66-81-9 | |
D(-)-Mannitol | Carl Roth | 4175.1 | |
D(+)-Glucose | Carl Roth | HN06.1 | |
DMEM (Dulbeccos Modified Eagle Medium) | PAN-Biotech | P04-03590 | |
DPBS (Dulbecco's Phosphate Buffer Saline) | PAN-Biotech | P04-36500 | |
Emission filter wheel (460/80, 535/70, 590/50, 642/80, 100%) | Leica | 11525480 | |
FBS (Fetal Bovine Serum) | PAN-Biotech | P30-3302 | |
Filter cube CYR71010 | Leica | 11525416 | |
FuGENE | Promega | E2691 | |
Glas Bottom Culture Dishes 35 mm | MatTek | P35G-0-10-C | |
HeLa cells | Leibniz Forschungsinstitut DSMZ | ACC 57 | Mammalian cervix carcinoma/ Obtained from Leibniz Forschungsinstitut DSMZ |
HEPES | Carl Roth | 9105.4 | |
ibidi µ-Disch 35 mm | ibidi | 81156 | |
KCl (Potassium chloride) | Carl Roth | 6781.1 | |
LAS X FRET Wizard | Leica | 11640862 | |
Light source LED8 | Leica | 11504256 | |
LRRC8A-mCerulean3 | N/A | N/A | König et al. 2019 |
LRRC8E-mVenus | N/A | N/A | König et al. 2019 |
Luna-II Automated Cell Counter | logos biosystems | L40002 | |
Luna-II Cell Counter Slides | logos biosystems | L12001 | |
MgCl2 (Magnesium chloride) | Carl Roth | KK36.1 | |
Microscope THUNDER Imager live cell | Leica | 11525681 | |
NaCl (Sodium chloride) | Carl Roth | 9263 | |
Objective HC PL APO 63x/1.40 OIL | Leica | 11506349 | |
Opti-Minimal Essential Medium (MEM) | gibco | 11058 | |
Osmometer OM807 | Vogel | V04807 | |
Penicillin Streptomycin (Pen Step) | gibco | 15070-063 | |
Trypan blue solution (0,4%) | Sigma | T8154 | |
Tumor necrosis factor (TNF)-a | Sigma-Aldrich | 94948-59-1 | |
Valve Controlled Gravity Perfusion System | ALA Scientific Instruments | VC3-4xG |
.